paper_id stringlengths 19 21 | paper_title stringlengths 8 170 | paper_abstract stringlengths 8 5.01k | paper_acceptance stringclasses 18
values | meta_review stringlengths 29 10k | label stringclasses 3
values | review_ids list | review_writers list | review_contents list | review_ratings list | review_confidences list | review_reply_tos list |
|---|---|---|---|---|---|---|---|---|---|---|---|
nips_2022_t6O08FxvtBY | Advancing Model Pruning via Bi-level Optimization | The deployment constraints in practical applications necessitate the pruning of large-scale deep learning models, i.e., promoting their weight sparsity. As illustrated by the Lottery Ticket Hypothesis (LTH), pruning also has the potential of improving their generalization ability. At the core of LTH, iterative magnitud... | Accept | The reviewers had significantly diverging opinions on this manuscript. The main issue under discussion was whether the framing of this paper as a lottery ticket work was correct, given that the main evaluations use no reinitialization or rewinding. On balance, I think that while one reviewer was very negative about the... | val | [
"k0_Hr_4oX9c",
"SU9wf_zsJj7",
"Q1S0FPLOen_",
"iqE3UuIVU6w",
"KV5uNe2P1f",
"w0uMl2b2CQu",
"j4zV7io0Kz5",
"7_xIW1Rvdi1",
"py64YZVA8fG",
"9mgpDJnpnHQ",
"xQKR7_vc4kZ",
"kvP_reBAY95",
"LgRy-aZggvz",
"ckz79KyIMNi",
"f5j1qts4oPL",
"1ClJcuzWCYS",
"56-6HcaPDxx",
"05nXc4nigBB",
"16FWk0si9c... | [
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author"... | [
" Thank you for posting [Response to authors](https://openreview.net/forum?id=t6O08FxvtBY¬eId=Q1S0FPLOen_). Please see our follow-up clarification and response below.\n\n**Q1: Since novelty of contributions was listed as one of the initial weaknesses — No, that wasn't in my review or my response.**\n\nA1: The co... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
7,
8,
2
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4,
5
] | [
"w0uMl2b2CQu",
"Q1S0FPLOen_",
"w0uMl2b2CQu",
"7_xIW1Rvdi1",
"7_xIW1Rvdi1",
"7_xIW1Rvdi1",
"py64YZVA8fG",
"kvP_reBAY95",
"LgRy-aZggvz",
"K-BISb3ND_a",
"OX68iAHfmv",
"D-YYStKy9FM",
"cQiBjp_bI0P",
"nips_2022_t6O08FxvtBY",
"cQiBjp_bI0P",
"05nXc4nigBB",
"05nXc4nigBB",
"jzC7z6ztw2o",
"... |
nips_2022__w-ivKc1cj | Learn what matters: cross-domain imitation learning with task-relevant embeddings | We study how an autonomous agent learns to perform a task from demonstrations in a different domain, such as a different environment or different agent. Such cross-domain imitation learning is required to, for example, train an artificial agent from demonstrations of a human expert. We propose a scalable framework that... | Accept | All three reviewers have elected to accept the paper, with two weak accepts and one accept. The reviews were thorough and demonstrated an understanding of the paper, and the authors have addressed many of the suggested edits.
I find figure 2 of the paper (comparison to XIRL on XMagical benchmark) compelling.
Recommend... | val | [
"mXIxZ5suybP",
"iyRUVFt7iW",
"fpFap6hXOUX",
"bA1NzvlyChe",
"N0jb191LUvL",
"S8sRJsCAxmF",
"DFOQYNOULHkn",
"SVXX1dLb-8T",
"AdbUDNl_GUb",
"FTaTgTMQwfi",
"Iwa3uyYtQnm",
"VX4wzQPy9Sc"
] | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for the thorough response and additional evaluations. I will be keeping the original score (7). ",
" Thanks, very thorough response. I don't know if you just made all the changes just because I requested them, or if you actually think it makes the paper better, but regardless, personally I do think th... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4
] | [
"N0jb191LUvL",
"S8sRJsCAxmF",
"nips_2022__w-ivKc1cj",
"nips_2022__w-ivKc1cj",
"VX4wzQPy9Sc",
"Iwa3uyYtQnm",
"Iwa3uyYtQnm",
"Iwa3uyYtQnm",
"FTaTgTMQwfi",
"nips_2022__w-ivKc1cj",
"nips_2022__w-ivKc1cj",
"nips_2022__w-ivKc1cj"
] |
nips_2022__VF5QKgXoqt | HumanLiker: A Human-like Object Detector to Model the Manual Labeling Process | Popular object detection models generate bounding boxes in a different way than we humans. As an example, modern detectors yield object box either upon the regression of its center and width/height (center-guided detector), or by grouping paired estimated corners (corner-guided detector). However, that is not the patte... | Accept | This paper received borderline reviews, with one review leaning negative. However, the reviewer acknowledged that their concerns have been addressed but did not update the rating.
The paper provides an interesting new take on object detection with strong empirical results. The concerns raised by reviewers were mainly... | train | [
"A34q6VAIT3",
"s-iJaWAViw5",
"1x60o2EMVmZ",
"1KTFiYY9uU",
"HjQJvOG6Tq",
"cQe9Q40DrUH",
"8iINixQygGc"
] | [
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear authors,\n\nThank you for the detailed responses which address my concerns.\n\nBest",
" Thank you very much for your careful review and valuable comments. We address your concerns as follows:\n\n1. This question is interesting. However, we believe your hypothesis that humans may deduce the top-left corner ... | [
-1,
-1,
-1,
-1,
5,
5,
4
] | [
-1,
-1,
-1,
-1,
5,
5,
5
] | [
"s-iJaWAViw5",
"8iINixQygGc",
"cQe9Q40DrUH",
"HjQJvOG6Tq",
"nips_2022__VF5QKgXoqt",
"nips_2022__VF5QKgXoqt",
"nips_2022__VF5QKgXoqt"
] |
nips_2022_Gpqqm4p91Ez | Towards Lightweight Black-Box Attack Against Deep Neural Networks | Black-box attacks can generate adversarial examples without accessing the parameters of target model, largely exacerbating the threats of deployed deep neural networks (DNNs). However, previous works state that black-box attacks fail to mislead target models when their training data and outputs are inaccessible. In thi... | Accept | The paper presents a new method for generating black-box attacks with very limited data, i.e., the no-box case. The attack is based on feature transformations and the paper proposes error transformers (ETF) to alleviate issues with approximation errors. The reviewers believe the paper is technically solid and raised is... | train | [
"czCF6MCk8PL",
"Habl8C7tkZR",
"om_ke3pRLwQ",
"2-AqQIZxYVv",
"Dv2wL6MXxhO",
"LWjDxYtXngZ",
"TlZIPoj8NnvM",
"kN7AQTRhHS0",
"tj2HEHBedvQ",
"Rb-9-tYThQP",
"K5IHiMFYZBl",
"h8UMB04qJa",
"iecSuyuVjp",
"-8oOWTl4aT",
"VJOfeStDB60",
"CCkcYNQjDA8",
"PIG1oEVpaOL",
"QRzmO9mqp2l",
"nAaY9LKs4PQ... | [
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"officia... | [
" Dear Reviewer LA1D:\n\nThank you again for your valuable comments and constructive suggestions on our work. Would you mind checking the response to confirm whether it addresses your questions/concerns?\n\nSince we list many points in the response and the window for discussion is closing, we want to summarize our ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
5,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
3,
3
] | [
"Hdlih3dePeW",
"LWjDxYtXngZ",
"nips_2022_Gpqqm4p91Ez",
"tp0TnbL1DHj",
"LWjDxYtXngZ",
"HgYG224Hn9J",
"kN7AQTRhHS0",
"CCkcYNQjDA8",
"Hdlih3dePeW",
"Hdlih3dePeW",
"Hdlih3dePeW",
"Hdlih3dePeW",
"-Jjj-KeaDj8",
"-Jjj-KeaDj8",
"-Jjj-KeaDj8",
"-Jjj-KeaDj8",
"tp0TnbL1DHj",
"tp0TnbL1DHj",
... |
nips_2022_BZ92dxDS3tO | OnePose++: Keypoint-Free One-Shot Object Pose Estimation without CAD Models | We propose a new method for object pose estimation without CAD models. The previous feature-matching-based method OnePose has shown promising results under a one-shot setting which eliminates the need for CAD models or object-specific training. However, OnePose relies on detecting repeatable image keypoints and is thus... | Accept | This paper originally received slightly positive reviews overall, except for one review, which was plenty of requests of specific clarifications and comments. Main issues regarded just the need of clarifying some parts of the method and put better in context of the state of the art and former evaluations. Unclear novel... | val | [
"tsi6AjpEVys",
"iYiTEpz2Bde",
"mIZZrjxryt",
"U3mMiBBhh3D",
"KHKY75jO7vu",
"k6hTyESR8f4",
"gvsHS3HJQdW",
"x4p0nN_K--z",
"2EYVgYc0E64",
"t9b298mKdGs",
"0UrqmCnxjU",
"WQc0ZgE92MW",
"kQVgUrUfsuR",
"ZphKQxqijC",
"-HwdPtmJMvw",
"8HujlfSeJqc",
"_tyeP4Sljo1y",
"i1x3fLn9X2O",
"L-BU2a6v9k9... | [
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_revi... | [
" Thank you very much for the detailed description of the relation to prior work.\n\n> Our idea is to directly disambiguate and augment features by encoding their spatial information and relations with others into features with the help of the attention module.\n\nThere is actually prior work on using 3D point rela... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
5,
3,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
4,
3
] | [
"2EYVgYc0E64",
"x4p0nN_K--z",
"t9b298mKdGs",
"0UrqmCnxjU",
"gvsHS3HJQdW",
"WQc0ZgE92MW",
"ZphKQxqijC",
"-HwdPtmJMvw",
"-HwdPtmJMvw",
"8HujlfSeJqc",
"_tyeP4Sljo1y",
"kQVgUrUfsuR",
"QEJXgCgfevt",
"i1x3fLn9X2O",
"Mt2SG4bBbry",
"_tyeP4Sljo1y",
"L-BU2a6v9k9",
"XUE5-0wKa9D",
"OSrNdlttd... |
nips_2022_rY2wXCSruO | DeepInteraction: 3D Object Detection via Modality Interaction | Existing top-performance 3D object detectors typically rely on the multi-modal fusion strategy. This design is however fundamentally restricted due to overlooking the modality-specific useful information and finally hampering the model performance. To address this limitation, in this work we introduce a novel modality ... | Accept | This work was overall positively technically evaluated with some concerns mainly related to limited experimental validation, the need to some additional justifications and explanation, and the missing computational cost analysis.
The provided rebuttal responded sufficiently well to these concerns and the overall evalua... | val | [
"aZXN_Ygp2S0",
"T2lWuWxWpGY",
"afMsveYEh7r",
"fERc0MnS7Dw",
"SCb34geObeO",
"DV9l9IR0_Nn",
"nttt4noTrg",
"7CQ3GATLyZn",
"I3ixA8LORn0"
] | [
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" The paper describes a 3D object detection architecture for 3D LiDAR point clouds. The question at hand is whether or not the authors violated double-blind reviewing.\n\nHere are the facts:\n - The authors reference their position and test set results on a public leaderboard.\n -The leaderboard is the nuScenes det... | [
-1,
-1,
-1,
-1,
-1,
-1,
6,
1,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4
] | [
"nips_2022_rY2wXCSruO",
"DV9l9IR0_Nn",
"nips_2022_rY2wXCSruO",
"I3ixA8LORn0",
"7CQ3GATLyZn",
"nttt4noTrg",
"nips_2022_rY2wXCSruO",
"nips_2022_rY2wXCSruO",
"nips_2022_rY2wXCSruO"
] |
nips_2022_Qb-AoSw4Jnm | MoVQ: Modulating Quantized Vectors for High-Fidelity Image Generation | Although two-stage Vector Quantized (VQ) generative models allow for synthesizing high-fidelity and high-resolution images, their quantization operator encodes similar patches within an image into the same index, resulting in a repeated artifact for similar adjacent regions using existing decoder architectures. To addr... | Accept | The three reviewers had significantly diverging final opinions (strong accept, borderline accept and weak reject). The authors addressed many of the concerns in their rebuttal. I read the paper carefully, and I agree with the concerns from one reviewer about why the improvements in stage-1 do not lead to significant im... | test | [
"WlETZSG0I8x",
"VLLP3zzno17",
"cHtiZr8dowOS",
"2yZd9pMfOHKZ",
"wCX7KXajPC1",
"l35Xob3uDP6",
"WlNdWjaJ6BG"
] | [
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear reviewers,\n\nWe first thank you again for your valuable comments and detailed suggestions. In the previous replies, we have tried our best to address your questions and revised the manuscript based on the suggestions.\n\nWe are looking forward to your reply to our responses, and we are open to any discussio... | [
-1,
-1,
-1,
-1,
7,
5,
4
] | [
-1,
-1,
-1,
-1,
4,
4,
5
] | [
"nips_2022_Qb-AoSw4Jnm",
"WlNdWjaJ6BG",
"l35Xob3uDP6",
"wCX7KXajPC1",
"nips_2022_Qb-AoSw4Jnm",
"nips_2022_Qb-AoSw4Jnm",
"nips_2022_Qb-AoSw4Jnm"
] |
nips_2022_A1yGs_SWiIi | TransTab: Learning Transferable Tabular Transformers Across Tables | Tabular data (or tables) are the most widely used data format in machine learning (ML). However, ML models often assume the table structure keeps fixed in training and testing. Before ML modeling, heavy data cleaning is required to merge disparate tables with different columns. This preprocessing often incurs significa... | Accept | This work introduces and evaluates a general scheme to feature-ize tabular data, and methods for (self-supervised) pre-training over the same, with a focus is on learning transferable representations.
Reviewers were unanimous that the approach proposed constitutes a flexible, practical approach that borrows and bring... | train | [
"c_QwlRb8o_A",
"r8pkZwHATj",
"A-fU7Ly1dRj",
"aLqS_gDTdo",
"GnGa9cc7wyy",
"xCtMBogBz4I",
"0kjRg9OQ2Ey",
"LnZ9YFu_mmi",
"bXi2zHlJhWY",
"e6ZzIaB8AkE"
] | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Most of my queries/doubts/concerns are answered and I also increased the rating to 7.",
" Thanks for addressing my questions and concerns and I am happy to raise my score to 7.",
" We thank Reviewer THBY for the helpful feedback. Besides our general response above, please see our specific response below.\n\n... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
7,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4
] | [
"aLqS_gDTdo",
"GnGa9cc7wyy",
"e6ZzIaB8AkE",
"bXi2zHlJhWY",
"xCtMBogBz4I",
"LnZ9YFu_mmi",
"nips_2022_A1yGs_SWiIi",
"nips_2022_A1yGs_SWiIi",
"nips_2022_A1yGs_SWiIi",
"nips_2022_A1yGs_SWiIi"
] |
nips_2022_1tnVNogPUz9 | Towards Efficient 3D Object Detection with Knowledge Distillation | Despite substantial progress in 3D object detection, advanced 3D detectors often suffer from heavy computation overheads. To this end, we explore the potential of knowledge distillation (KD) for developing efficient 3D object detectors, focusing on popular pillar- and voxel-based detectors. In the absence of well-devel... | Accept | In this paper, the authors propose a new method for knowledge distillation for 3D object detection in point cloud data. This problem is quite important for self-driving cars and 3D computer vision. The goal of their work is to compress models to achieve reasonable trade-offs in compute performance versus accuracy. The ... | train | [
"tHeuIjZSN7N",
"TQWDiCKieY",
"bLehisoipfG",
"CNNoeZ_Mj-e",
"QR4SHHlVVd",
"OxCqXnR6moZ",
"8J2JZeWqyF",
"nrJhpDI99ES",
"8QEArRFtRP5M",
"ckgnNe9eeau",
"42XGlIEtU9P",
"TwL3kRC6XhO",
"G8s6lR2jpYh",
"IkhsI_iIMZq",
"NpvwiowPx6C",
"yGFXqg8h8RQ",
"WPtMIpj8Xqx"
] | [
"author",
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for your constructive comments and suggestions. If you have other questions and concerns, please let us know and we are happy to further discuss. Thank you again for your time.",
" Thank you for your constructive comments and suggestions. If you have other questions and concerns, please let us know an... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
7,
8
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
5,
5,
2
] | [
"NpvwiowPx6C",
"yGFXqg8h8RQ",
"CNNoeZ_Mj-e",
"G8s6lR2jpYh",
"OxCqXnR6moZ",
"42XGlIEtU9P",
"NpvwiowPx6C",
"NpvwiowPx6C",
"NpvwiowPx6C",
"IkhsI_iIMZq",
"IkhsI_iIMZq",
"yGFXqg8h8RQ",
"WPtMIpj8Xqx",
"nips_2022_1tnVNogPUz9",
"nips_2022_1tnVNogPUz9",
"nips_2022_1tnVNogPUz9",
"nips_2022_1tn... |
nips_2022_Ag3ycrdh6n | Tensor Wheel Decomposition and Its Tensor Completion Application | Recently, tensor network (TN) decompositions have gained prominence in computer vision and contributed promising results to high-order data recovery tasks. However, current TN models are rather being developed towards more intricate structures to pursue incremental improvements, which instead leads to a dramatic increa... | Accept | Two reviewers consider that the proposed construction is clearly innovative.
and all reviewers consider that the contribution is useful to the tensor learning community.
The experiments show that the proposed method yields improved performance.
The three reviewers who participated in the discussion with the authors... | val | [
"PMQzpwvf5Vj",
"4dubWirRtz",
"8p9CnV5XMJX",
"VvMmtjyzHl",
"RUt2NBR_bs5",
"nkFATfltdMZ",
"8Br49Nwuyyr",
"8nsh909zVR",
"5ShbRcaseLs",
"Q36uXfk4GL2",
"Fut7h-oeN2N",
"achcnnzuIEf",
"uop69AIaaTc",
"oRLhC4-nR_",
"B0bOrAgVWKL",
"oGe28T4F2c",
"RtMu9CNvSb",
"uP_0-6b8xHT"
] | [
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We have detailedly responded to your comments and carefully addressed your concerns. Thanks again for making our results even stronger. Sincerely, we look forward to further communication with you!\n",
" Thanks for the reviewer's insightful and valuable comments to our work. We have posted all responses in the ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
4,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
4
] | [
"RtMu9CNvSb",
"RtMu9CNvSb",
"RUt2NBR_bs5",
"8Br49Nwuyyr",
"oRLhC4-nR_",
"nips_2022_Ag3ycrdh6n",
"8nsh909zVR",
"uop69AIaaTc",
"Q36uXfk4GL2",
"Fut7h-oeN2N",
"uP_0-6b8xHT",
"RtMu9CNvSb",
"oGe28T4F2c",
"B0bOrAgVWKL",
"nips_2022_Ag3ycrdh6n",
"nips_2022_Ag3ycrdh6n",
"nips_2022_Ag3ycrdh6n",... |
nips_2022_JGLW4DvX11F | Optimistic Tree Searches for Combinatorial Black-Box Optimization | The optimization of combinatorial black-box functions is pervasive in computer science and engineering. However, the combinatorial explosion of the search space and lack of natural ordering pose significant challenges for current techniques from a theoretical and practical perspective, and require new algorithmic ideas... | Accept | This paper proposes two methods for black box optimization of Lipschitz combinatorial binary functions. The reviewers agree that the paper is well written, the methods are sufficiently novel, and that the results are of interest to the NeurIPS community. The main drawback with the paper is that reviewer n1bW felt that ... | train | [
"fYutebcKKkK",
"v8KEiUH9TOn",
"ksvNtferRN",
"QGxr8kA6WU1",
"M0ghQXQ63M",
"Bg0E893mOR",
"0cOmCIejYB",
"Y91yYhD2-5zp",
"Ef5r10DluXk",
"j6GbWfY5j_y",
"Iyc6FPf1-i",
"W2D2bB2kGQQ",
"SXqhojOaJPB",
"vHpY-aULDH7",
"PJ672Qkdxk",
"VVyfDbYztk5"
] | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Finally, we would like to thank the reviewer for taking the time to read the rebuttal and we are happy that it helped to clarify some of its questions.",
" Thank you for pointing out the applicability of the approach.\nFor the indices, we agree that the large budget (considered \nin the current setting) contrib... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
7,
4,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
2,
4,
3
] | [
"ksvNtferRN",
"QGxr8kA6WU1",
"W2D2bB2kGQQ",
"Bg0E893mOR",
"0cOmCIejYB",
"Y91yYhD2-5zp",
"Ef5r10DluXk",
"j6GbWfY5j_y",
"VVyfDbYztk5",
"PJ672Qkdxk",
"vHpY-aULDH7",
"SXqhojOaJPB",
"nips_2022_JGLW4DvX11F",
"nips_2022_JGLW4DvX11F",
"nips_2022_JGLW4DvX11F",
"nips_2022_JGLW4DvX11F"
] |
nips_2022_GKfNB4BegL | Recurrent Video Restoration Transformer with Guided Deformable Attention | Video restoration aims at restoring multiple high-quality frames from multiple low-quality frames. Existing video restoration methods generally fall into two extreme cases, i.e., they either restore all frames in parallel or restore the video frame by frame in a recurrent way, which would result in different merits and... | Accept | The paper introduces a recurrent video restoration transformer with guided attention, which combines recurrent and parallel methods in some extent. All reviewers found that the proposed method is sound and that experiments are adequate to demonstrate the effectiveness of the proposed method. | train | [
"9RfLAMTp7KI",
"xBF1bnvxURG",
"y5FpW7w399k",
"YVqmTm2pTS",
"UcRo-K6kmQ",
"AIohC8Rt0LF",
"0rca7LYHs9h",
"7V7bRaikglL",
"2jogGt8dv29"
] | [
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for your responses. The authors address my concerns.",
" Thank the authors for their detailed responses. My concerns have been well addressed. I recommend acceptance of this work. ",
" I appreciate your responses. All of my concerns have been addressed by the response. After seeing the rebuttal and oth... | [
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
8
] | [
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
5
] | [
"0rca7LYHs9h",
"7V7bRaikglL",
"YVqmTm2pTS",
"2jogGt8dv29",
"7V7bRaikglL",
"0rca7LYHs9h",
"nips_2022_GKfNB4BegL",
"nips_2022_GKfNB4BegL",
"nips_2022_GKfNB4BegL"
] |
nips_2022_kCTZt0b9DQz | Prototypical VoteNet for Few-Shot 3D Point Cloud Object Detection | Most existing 3D point cloud object detection approaches heavily rely on large amounts of labeled training data. However, the labeling process is costly and time-consuming. This paper considers few-shot 3D point cloud object detection, where only a few annotated samples of novel classes are needed with abundant samples... | Accept | The paper received mixed reviews. Two reviewers were fairly positive, on the basis of of the novelty of the problem, the quality of the results, the introduction of datasets and benchmarks, and the proposed method. Although the method combines two existing solutions to point cloud detection and few-shot, these reviewer... | train | [
"du7vQwNT3bV",
"6Sv5xU7jwR",
"15c0W0nz_fw",
"c1FWpz0jfRi",
"cp6OeYvJOSs",
"ZrwNabRSX3h",
"Dp1o2o0djGI",
"j2MHVvt1DQk",
"Qr0SK31Gl6",
"MMSmLz1orKP",
"5BaOvaoCXr",
"wbtPtMOuEb2",
"IU383M4S6Jl",
"3CL_rVet-IP",
"DKMXV7m6Hsm",
"q0i9Wignlw0",
"cexKDfbqTwv",
"fTI8V3h_jf3",
"TOEAOs1OYrC"... | [
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear Reviewer ymyp,\n\nThank you so much for your time and efforts in assessing our paper. Hope our rebuttal has addressed your concerns. We are happy to discuss with you further if you still have other concerns. Thanks for helping improve our paper.",
" Thank you for the detailed and through response to my con... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
4,
7,
4
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
2,
4,
4,
5
] | [
"cexKDfbqTwv",
"DKMXV7m6Hsm",
"c1FWpz0jfRi",
"ZrwNabRSX3h",
"ZrwNabRSX3h",
"Dp1o2o0djGI",
"j2MHVvt1DQk",
"Qr0SK31Gl6",
"TOEAOs1OYrC",
"TOEAOs1OYrC",
"fTI8V3h_jf3",
"cexKDfbqTwv",
"cexKDfbqTwv",
"q0i9Wignlw0",
"q0i9Wignlw0",
"nips_2022_kCTZt0b9DQz",
"nips_2022_kCTZt0b9DQz",
"nips_20... |
nips_2022_dVXO3Orjmxk | Decoupling Classifier for Boosting Few-shot Object Detection and Instance Segmentation | This paper focus on few-shot object detection~(FSOD) and instance segmentation~(FSIS), which requires a model to quickly adapt to novel classes with a few labeled instances. The existing methods severely suffer from bias classification because of the missing label issue which naturally exists in an instance-level few-s... | Accept | **Summary**: This paper aims to address the missing label issue in few-shot object detection (FSOD) and instance segmentation (FSIS). In these tasks, some foreground examples (more specifically, classes) are not labeled in the training images, which causes classification bias for the conventional classification head. T... | train | [
"3dJRGRYjIf2",
"SDYgVSjO7Kz",
"73yc0_OcuT9",
"l3Y_GvHI0hg",
"kU5llO77TxE",
"HkWyDEvLtOv",
"LKO50POFLGu",
"cq0PzGZ2RzD",
"sjz1gANqVzk",
"Zmy0bUdNGO8",
"cFwD7idc_lu",
"2GBkvDgkclE",
"mLL6HxGSYaZ",
"HOSUAdkyu6J",
"uh1_KGH7ZkQ",
"SC2oWgFdn_i",
"Ya4tNviElV"
] | [
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear reviewer LjLJ,\n\nThank you for taking the time to review our work. We have provided corresponding responses, which we believe have covered your questions and concerns. We want to further discuss with you whether or not your concerns have been addressed. \n\nBest,",
" Dear reviewer 1weo,\n\nThank you for t... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
4,
3,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4,
5
] | [
"HOSUAdkyu6J",
"uh1_KGH7ZkQ",
"SC2oWgFdn_i",
"kU5llO77TxE",
"HkWyDEvLtOv",
"Ya4tNviElV",
"SC2oWgFdn_i",
"SC2oWgFdn_i",
"uh1_KGH7ZkQ",
"HOSUAdkyu6J",
"HOSUAdkyu6J",
"HOSUAdkyu6J",
"nips_2022_dVXO3Orjmxk",
"nips_2022_dVXO3Orjmxk",
"nips_2022_dVXO3Orjmxk",
"nips_2022_dVXO3Orjmxk",
"nips... |
nips_2022_BYLysbfdJOd | Planckian Jitter: countering the color-crippling effects of color jitter on self-supervised training | Several recent works on self-supervised learning are trained by mapping different augmentations of the same image to the same feature representation. The data augmentations used are of crucial importance to the quality of learned feature representations. In this paper, we analyze how the color jitter traditionally used... | Reject | Training representations in computer vision typically requires systematic augmentations of the input training set such as crops, reflections, translations including color jitter. As illumination invariance tends to be a desired property in visual object detection tasks, design of specific color augmentations is of grea... | train | [
"ODr6pL6Asm",
"m5VGggNCPYe",
"bV9UV5lmFwW",
"Yx4hzyrofw",
"3aIwtksGXpW",
"EabjOKKQ-Eg",
"eE8kdZBRC3K",
"9OHoJ9chkuR",
"wc5BD_KSj9T",
"fuIHpzs9vSr"
] | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you again for this discussion and for appreciating our experimental evaluation. We address your concerns below:\n\n>**Q3.1** I think the author may have misunderstood my question. My concern is that the experiment is misleading--it seems to suggest that the illuminant change proposed here is accurate and ca... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
7,
3
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
4
] | [
"m5VGggNCPYe",
"Yx4hzyrofw",
"Yx4hzyrofw",
"fuIHpzs9vSr",
"EabjOKKQ-Eg",
"wc5BD_KSj9T",
"9OHoJ9chkuR",
"nips_2022_BYLysbfdJOd",
"nips_2022_BYLysbfdJOd",
"nips_2022_BYLysbfdJOd"
] |
nips_2022_0ltDq6SjrfW | Efficient Knowledge Distillation from Model Checkpoints | Knowledge distillation is an effective approach to learn compact models (students) with the supervision of large and strong models (teachers). As empirically there exists a strong correlation between the performance of teacher and student models, it is commonly believed that a high performing teacher is preferred. Cons... | Accept | While this paper has 4 accept recommendations among the four reviewers, I have serious misgivings about the content of this paper. The main experimental insight, that a less trained teacher sometimes performs better, is already known and unsurprising. In fact it's the very point of KD that makes that result interesting... | train | [
"NJeUqWZaZn2",
"jxDM95r5b3U",
"zz2Mwj6-3Ei",
"Fid-rh3xJ4M",
"Oq8R2nYFYc",
"4L2qCccfUa",
"kjf3IHGqUqwd",
"XSgcqr0P_QG",
"kAJl-9lNR4f9",
"eGRSC3pLjp9Y",
"TWXE3Sy67cq",
"HyYOh1maYBa",
"FX7S9D9qyJ2",
"cfitxZnPw8g",
"IOxq5X5tgVP",
"oYAYWWHmKuC",
"XuRnXiClpmQ",
"AGRNAXj1rzh"
] | [
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks very much for your continued support. It encourages us a lot.\n\n(1) We sincerely accept the constructive suggestion of discussing contribution #2 more prominently. Some related study contents were put in the supplementary materials (such as, Section 2 and Section 4). We are considering to supplement a... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
6,
6,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
2
] | [
"Fid-rh3xJ4M",
"HyYOh1maYBa",
"4L2qCccfUa",
"eGRSC3pLjp9Y",
"XuRnXiClpmQ",
"oYAYWWHmKuC",
"IOxq5X5tgVP",
"kAJl-9lNR4f9",
"TWXE3Sy67cq",
"IOxq5X5tgVP",
"AGRNAXj1rzh",
"XuRnXiClpmQ",
"oYAYWWHmKuC",
"nips_2022_0ltDq6SjrfW",
"nips_2022_0ltDq6SjrfW",
"nips_2022_0ltDq6SjrfW",
"nips_2022_0l... |
nips_2022_wS23xAeKwSN | PolarMix: A General Data Augmentation Technique for LiDAR Point Clouds | LiDAR point clouds, which are usually scanned by rotating LiDAR sensors continuously, capture precise geometry of the surrounding environment and are crucial to many autonomous detection and navigation tasks. Though many 3D deep architectures have been developed, efficient collection and annotation of large amounts of ... | Accept | The proposed augmentation method for LIDAR Scans is to crop, cut, and mix two 3D scans at both scene-level and instance-level. The approach is not novel and a simple extension of the idea of the mix of 3D scenes and rotating bounding boxes. Another limitation is that the method cannot be applied to general 3D scenes. T... | test | [
"Fii_gLvQY0G",
"IFPM5-Ca2z0",
"SQJ6foE6CGX",
"4_wir9XSdJS",
"xdoU4mGVBM68",
"SFPolvdomAj",
"4SIJU6RtI_",
"LJiO8y_AcmN",
"9a_hAvigzUb",
"tIhr0vlkEIi",
"dRuVEqcSCh9",
"Z-gJyKUxw1_",
"IvWJFbiKQo",
"EPrBDQzp3qC",
"Sw7wN2jXODE",
"Hm4TOQrbev",
"RfG9A7s1Y6q"
] | [
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear Reviewer gjSm:\n\nWe thank you for the precious review time and valuable comments. We have provided corresponding responses and results, which we believe have covered your concerns. We hope to further discuss with you whether or not your concerns have been addressed. Please let us know if you still have any ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
7,
4,
6,
4
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
5,
4
] | [
"RfG9A7s1Y6q",
"Sw7wN2jXODE",
"IvWJFbiKQo",
"xdoU4mGVBM68",
"9a_hAvigzUb",
"4SIJU6RtI_",
"RfG9A7s1Y6q",
"Hm4TOQrbev",
"tIhr0vlkEIi",
"Sw7wN2jXODE",
"EPrBDQzp3qC",
"IvWJFbiKQo",
"nips_2022_wS23xAeKwSN",
"nips_2022_wS23xAeKwSN",
"nips_2022_wS23xAeKwSN",
"nips_2022_wS23xAeKwSN",
"nips_2... |
nips_2022_sde_7ZzGXOE | Is Out-of-Distribution Detection Learnable? | Supervised learning aims to train a classifier under the assumption that training and test data are from the same distribution. To ease the above assumption, researchers have studied a more realistic setting: out-of-distribution (OOD) detection, where test data may come from classes that are unknown during training (i.... | Accept | This paper studies generalization and learnability questions in the realm of out-of-distribution (OOD) detection. Specifically, it applies PAC learning to the theory of OOD detection. The contributions include new conceptual definitions of agnostic PAC learnability of OOD detection. Then, the authors argue for studying... | train | [
"76I8YaF2LK",
"MeSTGZRiBuE",
"vi1hXR6LBgx",
"Qvxyy5SnnsZ",
"_183zvTc8M",
"u3eKVG3Zi6N",
"2mD6XzVS0H",
"OaHRN-hvs5",
"5_9sfokjkK",
"_FHF-gIUWjt",
"U1IOGByRwM",
"-RvrWLeYR4Y",
"4Ep1FKIoPb_",
"TWaPSS0iHc"
] | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear Reviewer KK6q\n\nMany thanks for your kind support!\n\nDo you have more suggestions to improve the quality of our paper? We are glad to discuss our paper with you.\n\nBest,\n\nAuthors of Paper485",
" Thanks for the comprehensive response. I have to say sorry for giving an unfair score at first. Since I am ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
8,
8,
8
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
5,
1,
5
] | [
"MeSTGZRiBuE",
"Qvxyy5SnnsZ",
"4Ep1FKIoPb_",
"4Ep1FKIoPb_",
"-RvrWLeYR4Y",
"U1IOGByRwM",
"U1IOGByRwM",
"U1IOGByRwM",
"TWaPSS0iHc",
"TWaPSS0iHc",
"nips_2022_sde_7ZzGXOE",
"nips_2022_sde_7ZzGXOE",
"nips_2022_sde_7ZzGXOE",
"nips_2022_sde_7ZzGXOE"
] |
nips_2022_1qXIyIxLbEu | Let Images Give You More: Point Cloud Cross-Modal Training for Shape Analysis | Although recent point cloud analysis achieves impressive progress, the paradigm of representation learning from single modality gradually meets its bottleneck. In this work, we take a step towards more discriminative 3D point cloud representation using 2D images, which inherently contain richer appearance information, ... | Accept | The paper focuses on the problem of distilling semantic/representation knowledge from 2D images to help further enrich 3D point cloud representation. It received four detailed reviews and a healthy interaction between authors and reviewers ensued. In that back-and-forth, the reviewers clearly stated the weaknesses/issu... | train | [
"EOACATLBXgN",
"KkbhAvoHMe2",
"EF672tOSnLU",
"0QTivIw4ImB",
"UUKY-hZMfwX",
"CO6vuJkFW0c",
"ax5SmI3zAOG",
"8pOKe8Bx9h",
"sLr-9ZA7t4C",
"SEdO81Y4-Wc",
"5AlaQiPRJLP",
"WawDVXOhXj",
"EpuV85Z7bd_",
"cIITVr40Kcg",
"deziy14nKk3",
"UwzBykJRjbM"
] | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank for the additional experiment, after reading the author's response and other reviewers' comment, now I raised my score to 5",
" The response addressed my concerns. I will keep my Accept score.",
" We thank the reviewer’s beneficial suggestions. It seems the deadline of the author-reviewer discussion per... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
5,
5,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4,
5
] | [
"EF672tOSnLU",
"8pOKe8Bx9h",
"deziy14nKk3",
"UUKY-hZMfwX",
"CO6vuJkFW0c",
"ax5SmI3zAOG",
"SEdO81Y4-Wc",
"UwzBykJRjbM",
"deziy14nKk3",
"cIITVr40Kcg",
"EpuV85Z7bd_",
"nips_2022_1qXIyIxLbEu",
"nips_2022_1qXIyIxLbEu",
"nips_2022_1qXIyIxLbEu",
"nips_2022_1qXIyIxLbEu",
"nips_2022_1qXIyIxLbEu... |
nips_2022__efamP7PSjg | Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs | 3D-related inductive biases like translational invariance and rotational equivariance are indispensable to graph neural networks operating on 3D atomistic graphs such as molecules. Inspired by the success of Transformers in various domains, we study how to incorporate these inductive biases into Transformers. In this p... | Reject | This paper proposes Equiformer networks for predicting quantum properties based on 3D atomistic graphs. At the outset of the discussion period, the paper's scores were decidedly below borderline and the reviewers were concerned (i) that the methodological contribution of the paper was thin and (ii) about weaknesses in ... | train | [
"IpsdIJIgKn",
"r2qN0QJDDM4",
"Wf1q-dsBvW",
"hrA3aYa3xHM",
"sPYmkzBxBfQU",
"fKdUisDTftr",
"llZSxZZ8oeP",
"QWf8Wg0GVaW",
"AK4w1mboT9s",
"MVFUFaQyjWA",
"E0yy4S4UvvV",
"KW1QLwSknzT",
"q0uOPEKQwdN",
"PQXwcgO3jN7",
"RGrHCQCO5i-",
"u9wO3OXD1cM",
"WH-_ZGZth1-",
"zg-Vvns4vVA",
"4A-SQhT14A... | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"officia... | [
" We thank the reviewer for the response.\nSorry for having the tricks.",
" We thank the reviewer for the response and for acknowledging the thorough experimental evaluation.\nWe will improve the presentation of this paper and highlight the differences between existing works and this work.",
" \nI thank the aut... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
5,
5,
4
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
4
] | [
"Wf1q-dsBvW",
"hrA3aYa3xHM",
"zg-Vvns4vVA",
"AK4w1mboT9s",
"fKdUisDTftr",
"u9wO3OXD1cM",
"nips_2022__efamP7PSjg",
"fiaIYC_qd_b",
"MVFUFaQyjWA",
"E0yy4S4UvvV",
"KW1QLwSknzT",
"om_ojFrDpUD",
"PQXwcgO3jN7",
"RGrHCQCO5i-",
"DiVlrae79l",
"WH-_ZGZth1-",
"Zzah3p80u0x",
"5ajWccxDjL",
"ni... |
nips_2022_h10xdBrOxNI | Trap and Replace: Defending Backdoor Attacks by Trapping Them into an Easy-to-Replace Subnetwork | Deep neural networks (DNNs) are vulnerable to backdoor attacks. Previous works have shown it extremely challenging to unlearn the undesired backdoor behavior from the network, since the entire network can be affected by the backdoor samples. In this paper, we propose a brand-new backdoor defense strategy, which makes i... | Accept | The recommendation is based on the reviewers' comments, the area chair's personal evaluation, and the post-rebuttal discussion.
This paper proposed a new training method to defend against backdoor attacks. While all reviewers see merits in this paper, some discussions about (1) the practicality of the defense using c... | test | [
"N3GpycaCuB",
"ZH1AXcMiAFP",
"qw1flNvt2xK",
"d6L-N1JTWt0",
"UXXSix73dIY",
"lIm0CrIrAJ",
"lFoZX-akhv",
"AgAtGmCWZCw",
"KnV5Uy04wkn",
"dZ1rkH6OVZB",
"dqCQTiY9cLy",
"9hWtHlz6jI",
"WtXWsvAJhSi"
] | [
"author",
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for your detailed comments and insightful opinions! We think these constructive discussions are beneficial, to not only the authors, but also the entire community.\n\nWe agree that our experiments are conducted for the I.I.D. setting (i.e., \"case 3\") you mentioned. Below we would like to humbly defens... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
7,
4,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4,
5
] | [
"ZH1AXcMiAFP",
"d6L-N1JTWt0",
"UXXSix73dIY",
"UXXSix73dIY",
"lFoZX-akhv",
"WtXWsvAJhSi",
"9hWtHlz6jI",
"dqCQTiY9cLy",
"dZ1rkH6OVZB",
"nips_2022_h10xdBrOxNI",
"nips_2022_h10xdBrOxNI",
"nips_2022_h10xdBrOxNI",
"nips_2022_h10xdBrOxNI"
] |
nips_2022_UaXD4Al3mdb | Masked Autoencoders As Spatiotemporal Learners | This paper studies a conceptually simple extension of Masked Autoencoders (MAE) to spatiotemporal representation learning from videos. We randomly mask out spacetime patches in videos and learn an autoencoder to reconstruct them in pixels. Interestingly, we show that our MAE method can learn strong representations with... | Accept | This paper presents an interesting simple representation learning approach by extending masked autoencoders to videos. The reviewers have unanimously recognized the simplicity of approach, clarity of writing, and extensivity of experiments. Although there are some minor concerns about the novelty of the proposed method... | train | [
"_3WInM-97GL",
"zOCRC4EV8uO",
"vRHj2Qm932",
"YWKEG0wss11",
"2UTDxbBUpLi",
"XIL9JIVzB4",
"2hs8k3zfEo3",
"onDpLZY2bGK",
"crEUbe_7YSE",
"Q1kBXI-gHTU",
"K7umXF4QJhD",
"rygbNz6H_JG",
"Ae10ynOoywz",
"OZ8ixolEw55"
] | [
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We thank the reviewer for the feedback and are glad that most of the concerns are addressed. ",
" > Interesting. Does it work by removing the temporal tokenization (e.g. keep k_t=1), and predicting the time-slices based on the sampling stride. This removes one variable and could generalize better when evaluated... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
8,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
5
] | [
"YWKEG0wss11",
"XIL9JIVzB4",
"2UTDxbBUpLi",
"crEUbe_7YSE",
"K7umXF4QJhD",
"2hs8k3zfEo3",
"onDpLZY2bGK",
"Q1kBXI-gHTU",
"OZ8ixolEw55",
"Ae10ynOoywz",
"rygbNz6H_JG",
"nips_2022_UaXD4Al3mdb",
"nips_2022_UaXD4Al3mdb",
"nips_2022_UaXD4Al3mdb"
] |
nips_2022_V91cZ9i_sV3 | TOIST: Task Oriented Instance Segmentation Transformer with Noun-Pronoun Distillation | Current referring expression comprehension algorithms can effectively detect or segment objects indicated by nouns, but how to understand verb reference is still under-explored. As such, we study the challenging problem of task oriented detection, which aims to find objects that best afford an action indicated by verbs... | Accept | The paper proposes a Task-Oriented Instance Segmentation Transformer (TOIST) approach for finding objects that best afford a verb-indicated action, to handle the affordance recognition task. TOIST proposes two approaches of teacher-student knowledge distillation — it leverages the referring expression comprehension alg... | train | [
"jWp9MU300JF",
"O2Is6TpkmUy",
"aI-2tWYwIPz",
"lanLTb3UhJN",
"fA6HK_22u61",
"XNhYD-N1Yh73",
"IqeV4YEttY3",
"ilgz5ZsVJBb",
"vBvi2Y3zY5r",
"5js2NmH6XOL",
"3Oy28XfpTHa",
"ilZ3rhhFsnr",
"Fs5UsScafX3",
"PDgf8lpXW4",
"htj_uPRyp7G",
"jqXqWHtwgFj"
] | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for R#mBEm's feedback.\n\nFirstly, TOIST and MDETR+GGNN employ the same image encoder and text encoder. And the performance of MDETR+GGNN could derive from the pre-training process. The experimental results are shown below.\n\nTable 5: Comparison of different methods with the same backbone.\n| ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
5,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
3,
4
] | [
"O2Is6TpkmUy",
"5js2NmH6XOL",
"jqXqWHtwgFj",
"htj_uPRyp7G",
"PDgf8lpXW4",
"PDgf8lpXW4",
"jqXqWHtwgFj",
"jqXqWHtwgFj",
"jqXqWHtwgFj",
"htj_uPRyp7G",
"htj_uPRyp7G",
"PDgf8lpXW4",
"PDgf8lpXW4",
"nips_2022_V91cZ9i_sV3",
"nips_2022_V91cZ9i_sV3",
"nips_2022_V91cZ9i_sV3"
] |
nips_2022_QfI_usBXNCM | Cross-Image Context for Single Image Inpainting | Visual context is of crucial importance for image inpainting. The contextual information captures the appearance and semantic correlation between the image regions, helping to propagate the information of the complete regions for reasoning the content of the corrupted regions. Many inpainting methods compute the visual... | Accept | The paper discusses how to use external information for inpainting. Reviewers appreciated the idea but raised concerns regarding limited novelty, use of the proposed method for inpainting, baselines being evaluated incorrectly, and missing ablations. The rebuttal was able to address most of the concerns and reviewers r... | train | [
"3cChXTPwLTf",
"fORAdyxYPf",
"sogNU0IDIVx",
"r0C6XRdGd9u",
"0EORZ7yIbfO",
"yduObXg9QjX",
"4NxQ89GRH13",
"23PdKDoPI-",
"ZcKljQmMJfQ",
"DGAbWT1Hc-m",
"WmdxJFo8zCH",
"XFPUgEQbFj7",
"thgao51_GLk",
"nyVyW5uJ_Fb",
"M4r6m263It",
"n9R0tuK4p-L",
"kKTX5jqBzJM",
"IGCEEXu0h5J",
"Rk8Pv9so5K",... | [
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official... | [
" Dear Reviewer 3YLw,\n\nThanks for your review again. As the deadline for the authors' response is approaching, we sincerely request your comment on our primary response. This will definitely give us a valuable chance to address the questions unsolved.\n\nBest,\n\nAuthors of Paper ID 465",
" Dear Reviewer 8FQc,\... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
5,
5,
8,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
4,
4
] | [
"ADj0bvZt3Zt",
"_sTOGyFZGhO",
"IGCEEXu0h5J",
"Rk8Pv9so5K",
"yduObXg9QjX",
"w3Tk9zkOuG",
"ADj0bvZt3Zt",
"w3Tk9zkOuG",
"_sTOGyFZGhO",
"Rk8Pv9so5K",
"IGCEEXu0h5J",
"ADj0bvZt3Zt",
"w3Tk9zkOuG",
"_sTOGyFZGhO",
"Rk8Pv9so5K",
"IGCEEXu0h5J",
"nips_2022_QfI_usBXNCM",
"nips_2022_QfI_usBXNCM"... |
nips_2022_-T5seeOMnM5 | Natural Color Fool: Towards Boosting Black-box Unrestricted Attacks | Unrestricted color attacks, which manipulate semantically meaningful color of an image, have shown their stealthiness and success in fooling both human eyes and deep neural networks. However, current works usually sacrifice the flexibility of the uncontrolled setting to ensure the naturalness of adversarial examples. A... | Accept | The proposed approach exploits the color distribution of semantic classes, thus improving the flexibility of the current unrestricted color attack. This method generates novel transferrable adversarial attacks. The authors conducted extensive experiments on wide variety of network architectures. A significant improveme... | train | [
"34h4RvRCvGx",
"PRvhq7AAMjU",
"HwMVOXPDUmQ",
"XFETUyxz_h0",
"ZVl0yksel6o",
"_iIaOOuYtLJ",
"fcR3GH2yWQ",
"tXZFUPXyW-z",
"FqDvb3nrt8f",
"1kqrnwlSY9c",
"y5YgYoPbv-0",
"B7zFhfwP4-k",
"rwiYWht9sx",
"Fo-EqTJUb0"
] | [
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Authors have convincingly addressed my concerns and I am willing to increase the score.",
" Thanks for addressing my concerns. I am willing to increase the score. ",
" Thank you a lot for the reply, which gives us the opportunity to address your concerns more clearly.\n\n_**Q4 What makes your attacks achieve ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
4
] | [
"fcR3GH2yWQ",
"HwMVOXPDUmQ",
"XFETUyxz_h0",
"FqDvb3nrt8f",
"_iIaOOuYtLJ",
"B7zFhfwP4-k",
"tXZFUPXyW-z",
"rwiYWht9sx",
"Fo-EqTJUb0",
"y5YgYoPbv-0",
"nips_2022_-T5seeOMnM5",
"nips_2022_-T5seeOMnM5",
"nips_2022_-T5seeOMnM5",
"nips_2022_-T5seeOMnM5"
] |
nips_2022_lXUp6skJ7r | Adversarial Style Augmentation for Domain Generalized Urban-Scene Segmentation | In this paper, we consider the problem of domain generalization in semantic segmentation, which aims to learn a robust model using only labeled synthetic (source) data. The model is expected to perform well on unseen real (target) domains. Our study finds that the image style variation can largely influence the model's... | Accept | Simple and practical way to do better at domain generalization when it comes to semantic segmentation. AdvStyle can generate images that are hard during training, and prevent the model from overfitting on the source domain. Given that it works well, is relatively simple to implement and conceptually sound, I think it w... | test | [
"FSNE6R-XGqZ",
"A8QcUInNalG",
"0atXthd0h2v",
"DlaPVj6PPj",
"b8OTPiql6II",
"A5c9qzI86Bx",
"PbdBbJQW1iJ",
"L3FZNv2-6Mh",
"dPo5SnFNnWX",
"I7SU4XUPODN",
"gdzGXPIFILJ",
"gRcSA0hKXc",
"5qILJQ3D6eH",
"acXP9BAaRk"
] | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear Reviewer NDMW,\n\nThanks for your kind reply. We are delighted that you appreciate our response. We definitely will add all these additional experiments into the final version. We also fully agree with you that the multi-resolution spatial pyramid strategy can bring further improvement and would like to leav... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
3,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4
] | [
"A8QcUInNalG",
"gdzGXPIFILJ",
"acXP9BAaRk",
"5qILJQ3D6eH",
"gRcSA0hKXc",
"acXP9BAaRk",
"acXP9BAaRk",
"5qILJQ3D6eH",
"5qILJQ3D6eH",
"5qILJQ3D6eH",
"gRcSA0hKXc",
"nips_2022_lXUp6skJ7r",
"nips_2022_lXUp6skJ7r",
"nips_2022_lXUp6skJ7r"
] |
nips_2022__yEcbgIT68e | HSurf-Net: Normal Estimation for 3D Point Clouds by Learning Hyper Surfaces | We propose a novel normal estimation method called HSurf-Net, which can accurately predict normals from point clouds with noise and density variations. Previous methods focus on learning point weights to fit neighborhoods into a geometric surface approximated by a polynomial function with a predefined order, based on w... | Accept | This paper proposes an approach to fit implicit surfaces for surface normal estimation. Reviewers unanimously agree on its novelty and performance. AC hence recommends acceptance. | train | [
"oij664rsvQ",
"xz8TPT1Mmq",
"o5IMbK_7eyY",
"YyNK6gq_lV",
"jewHO9_HSG",
"oY4qrJu4DQw",
"bHr9OEliqy",
"XmWJKSzw9NI",
"17OSHflLoq_",
"hg2AXbfYUhY",
"Afp9ZqtBk_0",
"oE0n6gkWtJn",
"fp9uZrBsh5W",
"inJBplKsrkk",
"zYiC2E_Ncws",
"LIuBDxA9Ek",
"8CIYA-yc_1Q",
"B2jf6KM7boa"
] | [
"official_reviewer",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Additional experiments are always welcome, but just to clarify: I would condition acceptance on the changes I described above regardless of the outcome of such an experiment. That an explicit fitting with the approach described in Eq. 5 performs worse is not an indication that the network actually approximates Eq... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
5,
5,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4,
5
] | [
"xz8TPT1Mmq",
"o5IMbK_7eyY",
"17OSHflLoq_",
"jewHO9_HSG",
"oY4qrJu4DQw",
"Afp9ZqtBk_0",
"fp9uZrBsh5W",
"inJBplKsrkk",
"hg2AXbfYUhY",
"oE0n6gkWtJn",
"B2jf6KM7boa",
"8CIYA-yc_1Q",
"LIuBDxA9Ek",
"zYiC2E_Ncws",
"nips_2022__yEcbgIT68e",
"nips_2022__yEcbgIT68e",
"nips_2022__yEcbgIT68e",
... |
nips_2022_gkQkZy-pRik | MetaMask: Revisiting Dimensional Confounder for Self-Supervised Learning | As a successful approach to self-supervised learning, contrastive learning aims to learn invariant information shared among distortions of the input sample. While contrastive learning has yielded continuous advancements in sampling strategy and architecture design, it still remains two persistent defects: the interfere... | Accept | This paper starts with the experimentally findings that the interference of task irrelevant information and the disadvantages of sample inefficiency in contrastive learning appear due to dimensional redundancy and dimensional confounder. Based on these experimental findings, the authors propose a dimensional mask learn... | train | [
"Qo2HRTixOFi",
"jQG_RzcHBV",
"b46CeofYmCz",
"yBfEe6VQUjn",
"YZz1g2rHMEd",
"hRcb1F53F0k",
"kl1kw9AW-t2",
"Sn1bLiONFb",
"t9vrMbI192F",
"B8zn0d_0Csm",
"WRv2OU_GaDF",
"tEppK996_3R",
"CMHTpKYP5Ik",
"lSDlH6ZH5-I",
"b15bU2sH65o",
"wNLfpPMduBx",
"sNFrzDtmfPY",
"6DPAbK2hj6w",
"jsfIm5ch8xa... | [
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
... | [
" Since the discussion phase is closing soon, this response could be our last chance to discuss with the reviewer. We would be grateful for the careful review and constructive suggestions of the reviewer. We hope our rebuttals may make our intuition behind this paper more understandable and clearer. Following the r... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
6,
6,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
4,
4
] | [
"mhcrWMxos_y",
"Zf9nhuzfVu",
"H41PzShwI9",
"YZz1g2rHMEd",
"jsfIm5ch8xa",
"nips_2022_gkQkZy-pRik",
"mhcrWMxos_y",
"mhcrWMxos_y",
"mhcrWMxos_y",
"Zf9nhuzfVu",
"Zf9nhuzfVu",
"Zf9nhuzfVu",
"H41PzShwI9",
"H41PzShwI9",
"H41PzShwI9",
"H41PzShwI9",
"jsfIm5ch8xa",
"jsfIm5ch8xa",
"nips_202... |
nips_2022_vKBdabh_WV | Meta Optimal Transport | We study the use of amortized optimization to predict optimal transport (OT) maps from the input measures, which we call Meta OT. This helps repeatedly solve similar OT problems between different measures by leveraging the knowledge and information present from past problems to rapidly predict and solve new problems. O... | Reject | This paper proposes an amortized optimization approach for predicting optimal transport (OT) maps. Three reviewers found that the proposed method is interesting. However, some concerns raised by another reviewer on the improvement of the computational efficiency and the generality of the proposed method were raised:
... | train | [
"TEFRc700U-k",
"46SihsL-I9",
"yi1jfC2PxvY",
"t7UlgTjnK1",
"7FeKRZ80NAH",
"b-KeH74gJgZ",
"gBT7GILDQ1j",
"hQ6mC4_EhjQ",
"MJOZ0gJCMXU",
"iHGUYC-zNvY",
"qZ445NM_Fdt",
"eAep-RDgLYIM",
"kGrqHEcmTW",
"GOB1t5NIqK",
"7Di0JvA6YeV"
] | [
"official_reviewer",
"author",
"official_reviewer",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" This paper considers learning a meta-model to predict the solution to the OT problem. The idea is novel and the paper is well-written.\n\nCompared to other OT methods that run from scratch, it can save the time of learning OT potential for a new task. However, the experiments in this paper are only 2 or 3 dimensi... | [
3,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
6,
7
] | [
4,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4
] | [
"nips_2022_vKBdabh_WV",
"yi1jfC2PxvY",
"MJOZ0gJCMXU",
"GOB1t5NIqK",
"qZ445NM_Fdt",
"gBT7GILDQ1j",
"hQ6mC4_EhjQ",
"7Di0JvA6YeV",
"iHGUYC-zNvY",
"GOB1t5NIqK",
"kGrqHEcmTW",
"nips_2022_vKBdabh_WV",
"nips_2022_vKBdabh_WV",
"nips_2022_vKBdabh_WV",
"nips_2022_vKBdabh_WV"
] |
nips_2022_5VCT-DptDTs | Heterogeneous Skill Learning for Multi-agent Tasks | Heterogeneous behaviours are widespread in many multi-agent tasks, which have not been paid much attention in the community of multi-agent reinforcement learning. It would be a key factor for improving the learning performance to efficiently characterize and automatically find heterogeneous behaviours. In this paper, w... | Accept | The reviewers have largely agreed upon the value of the paper's concept (heterogeneous skills for MARL) and appreciated its impressive experimental gains on a range of environments. Each reviewer pointed out unique areas for improvement: citations to classical work, precision in derivations around conditional entropy a... | train | [
"i_tW9DWucI",
"4ROKEFRvQwZ",
"aLsJPqd3Y-X",
"8ahvVDHBfa-",
"eCwPkBDF5lo",
"p2Z3N-zntt",
"J0W62fywyJc",
"g00EqJSGE7",
"bZehBivUbfE",
"M2sQe29rYC",
"Pvt8gSVFZ-",
"dNOBm-1eKFv",
"EPlkwVwUcmz",
"O7IxswKA8zs"
] | [
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you very much for the inspiring and insightful comments which really helped us to improve our work!\nWe have submitted a revised paper. We hope that we have addressed all your concerns.\nPlease let me know if you have any other issues.",
" Thank you for the response. \nTotally, unclear points for me are c... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
4,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
3
] | [
"8ahvVDHBfa-",
"g00EqJSGE7",
"8ahvVDHBfa-",
"p2Z3N-zntt",
"p2Z3N-zntt",
"J0W62fywyJc",
"EPlkwVwUcmz",
"bZehBivUbfE",
"O7IxswKA8zs",
"Pvt8gSVFZ-",
"dNOBm-1eKFv",
"nips_2022_5VCT-DptDTs",
"nips_2022_5VCT-DptDTs",
"nips_2022_5VCT-DptDTs"
] |
nips_2022_RF5Lb6NaZp | End-to-End Learning to Index and Search in Large Output Spaces | Extreme multi-label classification (XMC) is a popular framework for solving many real-world problems that require accurate prediction from a very large number of potential output choices. A popular approach for dealing with the large label space is to arrange the labels into a shallow tree-based index and then learn an... | Accept | The paper considers extreme multilabel classification (XMC) and proposes a two-stage retrieval and classification model which replaces the usual initial hard-partitioning with a soft learnable partitioning. The reviewers concur that the end-to-end methodology for jointly training the representation, indexing, classific... | train | [
"OCtOJ1Y0_6",
"V1ibGOOn9DH",
"MDJHNo2vuLm",
"6P9ZIFPYiTlM",
"MXXeohPFqN",
"1YDUkPQCmKQ",
"d-W76dvQ7X0",
"xcnxoZ0hXn_",
"dtmHfgmFhBL",
"OzhxGuCHy_",
"SiPq9kAfKVS",
"LUi3e6XOQo",
"TmFn_GKaGA_",
"ArBzhqlbSrd"
] | [
"author",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We're happy to know that our response helped in addressing your concerns. Thanks again for the helpful feedback and upgrading the score!",
" The author-rebuttal phase closes today. Please acknowledge the author rebuttal and state if your position has changed. Thanks!",
" Thanks for the authors' response and i... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
6,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4
] | [
"MDJHNo2vuLm",
"TmFn_GKaGA_",
"SiPq9kAfKVS",
"MXXeohPFqN",
"1YDUkPQCmKQ",
"d-W76dvQ7X0",
"xcnxoZ0hXn_",
"dtmHfgmFhBL",
"ArBzhqlbSrd",
"TmFn_GKaGA_",
"LUi3e6XOQo",
"nips_2022_RF5Lb6NaZp",
"nips_2022_RF5Lb6NaZp",
"nips_2022_RF5Lb6NaZp"
] |
nips_2022_oWx_9VJgyV7 | SNAKE: Shape-aware Neural 3D Keypoint Field | Detecting 3D keypoints from point clouds is important for shape reconstruction, while this work investigates the dual question: can shape reconstruction benefit 3D keypoint detection? Existing methods either seek salient features according to statistics of different orders or learn to predict keypoints that are invaria... | Accept | Post-rebuttal, the paper had split reviews, with three reviewers in favor of acceptance (6, 6, 5 but noted as a 6 in the final comment from 2TwX) and one reviewer strongly arguing for rejection (3). The AC examined the reviews, the paper, and the discussion, and is inclined to accept the paper. The AC is persuaded by t... | train | [
"jgwi7mxF2Bk",
"0DeLpyqBYF",
"piE8lF-Ei3s",
"Ax7RUIPMpmp",
"7IyN4BNcdYP",
"ewce3jLLMX",
"pt2FaIMXNLkF",
"SWUOBwOpXDT",
"fefSnZhYs72",
"hwVFOCsLiENd",
"jabu1QKCRRz",
"OI37yz4nDUo",
"C-c4H4cWr2x",
"dcARkPno-5",
"WFiXaP9Q0Xz",
"PDx5_h8F0Ns"
] | [
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Hi Reviewers,\n\nThe discussion period is closing soon. Please take a look at the responses from the authors. If you have further questions, please ask them now, since the authors will be unable to respond soon. It's substantially more productive, effective, and reasonable to have a quick back-and-forth with auth... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
5,
3
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
5,
3,
5
] | [
"nips_2022_oWx_9VJgyV7",
"WFiXaP9Q0Xz",
"dcARkPno-5",
"7IyN4BNcdYP",
"SWUOBwOpXDT",
"PDx5_h8F0Ns",
"PDx5_h8F0Ns",
"PDx5_h8F0Ns",
"WFiXaP9Q0Xz",
"dcARkPno-5",
"OI37yz4nDUo",
"C-c4H4cWr2x",
"nips_2022_oWx_9VJgyV7",
"nips_2022_oWx_9VJgyV7",
"nips_2022_oWx_9VJgyV7",
"nips_2022_oWx_9VJgyV7"... |
nips_2022_m7CmxlpHTiu | Self-Supervised Aggregation of Diverse Experts for Test-Agnostic Long-Tailed Recognition | Existing long-tailed recognition methods, aiming to train class-balanced models from long-tailed data, generally assume the models would be evaluated on the uniform test class distribution. However, practical test class distributions often violate this assumption (e.g., being either long-tailed or even inversely long-t... | Accept | The reviewers agreed the paper provides some nice insights into tackling the difficult and under-explored problem of test-agnostic long-tailed recognition. The reviewers appreciated the thorough experiments and ablations provided. The author response sufficiently addressed the key concerns the reviewers had. | train | [
"NcbgvC_6Kr-",
"bRCnqyrFlbo",
"zV3mhhcxq37",
"srMXzOosNBf",
"UBvffFZIxJxK",
"5IcsTqXEDK-",
"EzZ4q0kP8F",
"dRrZDKkq8Gl",
"7pvsNHBytc",
"GpK7cIjIc9",
"jN63a45nMz9",
"OQJbmFlT1DJ",
"CjoqrvlidJx",
"hmtLGfyuyP-"
] | [
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We sincerely appreciate your time and effort in reviewing our paper. Based on your constructive suggestions, the work has become more solid and thorough now.",
" Thank the authors for addressing some of my concerns. To reflect the authors' efforts, I will change my score to 5.",
" Thanks very much for the res... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
7,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
5,
5
] | [
"bRCnqyrFlbo",
"dRrZDKkq8Gl",
"srMXzOosNBf",
"GpK7cIjIc9",
"nips_2022_m7CmxlpHTiu",
"hmtLGfyuyP-",
"hmtLGfyuyP-",
"hmtLGfyuyP-",
"OQJbmFlT1DJ",
"OQJbmFlT1DJ",
"CjoqrvlidJx",
"nips_2022_m7CmxlpHTiu",
"nips_2022_m7CmxlpHTiu",
"nips_2022_m7CmxlpHTiu"
] |
nips_2022_A6O79ipjlJC | A Novel Matrix-Encoding Method for Privacy-Preserving Neural Networks (Inference) | In this work, we present a novel matrix-encoding method that is particularly convenient for neural networks to make predictions in a privacy-preserving manner using homomorphic encryption. Based on this encoding method, we implement a convolutional neural network for handwritten image classification over encryption. ... | Reject | The reviewers were unanimous in their recommendation to reject the paper. The authors' responded to the reviews but recognized the limitation of their submission, particularly in terms of missing comparisons to related work.
I want to take this opportunity to address the author who wrote in their rebuttal:
*"I will ... | train | [
"v0V4gEkiRsh",
"zrDqENROuWF",
"6WD5lYSvOhC",
"IWIlYkbJ58",
"hY50ZRw-qLP",
"4oDwB3PiUJ0",
"6sdM6jZfx_k",
"K5yLl2hdMSy",
"GTvm9knzZfV",
"0ZlkDuOCWCy",
"LqA5Eqf37kb",
"zOvaqO_U-LJ"
] | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" I have noticed that this paper lacks a comparison with previous methods. \n\nAnd I am very grateful for the time you and other reviewers spent reading my work.\n\nI hope this work doesn't waste your time.",
" I don't think I have read the paper on GAZELLE or the one about MiniONN (both look familiar to me, thou... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
2,
3,
3,
3
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
4
] | [
"6WD5lYSvOhC",
"IWIlYkbJ58",
"6sdM6jZfx_k",
"K5yLl2hdMSy",
"zOvaqO_U-LJ",
"LqA5Eqf37kb",
"0ZlkDuOCWCy",
"GTvm9knzZfV",
"nips_2022_A6O79ipjlJC",
"nips_2022_A6O79ipjlJC",
"nips_2022_A6O79ipjlJC",
"nips_2022_A6O79ipjlJC"
] |
nips_2022_gwsnBjNcVEe | Phase Transition from Clean Training to Adversarial Training | Adversarial training is one important algorithm to achieve robust machine learning models. However, numerous empirical results show a great performance degradation from clean training to adversarial training (e.g., 90+\% vs 67\% testing accuracy on CIFAR-10 dataset), which does not match the theoretical guarantee deliv... | Accept | Reviewers all agree that the theory in the paper is interesting and that it helps us understand the robustness accuracy tradeoff. Several reviewers raise the issue that they are unsure about how “phase transition” is defined in this article, and whether the observed behavior is indeed a phase transition in a typical ... | train | [
"wdsxc-gB-By",
"YBJJSeYnDB",
"IJrEn1DreGV",
"tlKR0KF0-RE",
"Y5Sy_J0Ii44",
"aGgh0fOA1vi",
"edWJc3Xl6II",
"lFFZlA4bGdv",
"lsa5yYLtmk"
] | [
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for reviewing our paper! Below are some responses to your questions and comments:\n\n1. Q1, $\\varepsilon$ and $\\epsilon$: These two are different notations. The former notation $\\varepsilon$ is the noise term in the response $y$, which has nothing to do with the attack. The latter notation $\\epsilon... | [
-1,
-1,
-1,
-1,
-1,
4,
6,
4,
3
] | [
-1,
-1,
-1,
-1,
-1,
3,
4,
3,
3
] | [
"lsa5yYLtmk",
"lFFZlA4bGdv",
"edWJc3Xl6II",
"aGgh0fOA1vi",
"nips_2022_gwsnBjNcVEe",
"nips_2022_gwsnBjNcVEe",
"nips_2022_gwsnBjNcVEe",
"nips_2022_gwsnBjNcVEe",
"nips_2022_gwsnBjNcVEe"
] |
nips_2022_G6cJsOOx2R3 | On Enforcing Better Conditioned Meta-Learning for Rapid Few-Shot Adaptation | Inspired by the concept of preconditioning, we propose a novel method to increase adaptation speed for gradient-based meta-learning methods without incurring extra parameters. We demonstrate that recasting the optimisation problem to a non-linear least-squares formulation provides a principled way to actively enforce a... | Accept | This paper was quite well received by reviewers, with scores of 5, 6, 6, 8.
Reviewers felt the paper was well written, clear and expressed an interesting core idea.
Experimental results compare against MAML and show clear improvements.
The key idea here was inspired by preconditioning, and the method here aims to incr... | train | [
"56XVBG8hplI",
"-Opey45F2n7",
"8-G7uh-F55d",
"AIjNkHawo-w",
"sjDLDAgOCnD",
"8t1wvhkn6Mx",
"b8QBMAd12iZ",
"Z-TfgTGxfWI",
"jCArqoIB-O0",
"a2Aa0bR8M_I",
"bRsV3yi02it",
"ZtKem4aEMSX",
"Y6DrS_4etf_",
"hnnjHQZRJCB",
"lAP9NWvj-P"
] | [
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for your response. I do appreciate it. My concerns are partially resolved and I will increase my score from 4 to 5, but I still think this paper is at the borderline for NeurIPS. I strongly encourage the authors to perform the apples-to-apple comparison against other recent methods and present the results ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
8,
6,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
3,
4
] | [
"a2Aa0bR8M_I",
"lAP9NWvj-P",
"AIjNkHawo-w",
"b8QBMAd12iZ",
"8t1wvhkn6Mx",
"lAP9NWvj-P",
"Z-TfgTGxfWI",
"hnnjHQZRJCB",
"Y6DrS_4etf_",
"bRsV3yi02it",
"ZtKem4aEMSX",
"nips_2022_G6cJsOOx2R3",
"nips_2022_G6cJsOOx2R3",
"nips_2022_G6cJsOOx2R3",
"nips_2022_G6cJsOOx2R3"
] |
nips_2022_VhgC3SMTiy | Audio-Driven Co-Speech Gesture Video Generation | Co-speech gesture is crucial for human-machine interaction and digital entertainment. While previous works mostly map speech audio to human skeletons (e.g., 2D keypoints), directly generating speakers' gestures in the image domain remains unsolved. In this work, we formally define and study this challenging problem of ... | Accept | This paper enjoyed a reasonable interaction between the authors and the reviewers, with the authors addressing the reviewers' concerns about the novelty of the proposed method, its specificity to the "talking head" scenario, the fact that the model is used in a speaker-dependent fashion, and some concerns about specifi... | train | [
"wYOgBQJ_gfC",
"hAWsVGewjz-",
"jD1b31Ixh4C",
"a-JeA-_0dev",
"CnhVJ2vPt78",
"6UyygMWoxqG",
"t1NSaIj9XBV",
"Oq-Aub4X3IN",
"phl6UWx9FvV",
"Il19lzvNRefs",
"Vei3dn-EmSV",
"aMGqLSOJUN-",
"aDTH8wQA0b9",
"34wu6JnMnWD",
"7MHxmiDDi7k",
"ArVUpvZA8l",
"VFOyRIY4lYJ",
"xltQySJn4d_",
"IzV7UiQ3W... | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_... | [
" Dear Reviewer WZ3S:\n\nWe are delighted to hear that your concerns are addressed! Many thanks again for your very constructive comments, which have helped us improve the quality of the paper significantly.\n\nBest,\n\nPaper 365 Authors.\n\n",
" We sincerely thank the reviewer for the additional feedbacks. We us... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
7,
6,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
4,
3
] | [
"jD1b31Ixh4C",
"a-JeA-_0dev",
"t1NSaIj9XBV",
"phl6UWx9FvV",
"uSSf9VZAo-k",
"BSAJ6QD7gaS",
"N6A451n51Ed",
"aMGqLSOJUN-",
"uSSf9VZAo-k",
"BSAJ6QD7gaS",
"aMGqLSOJUN-",
"ArVUpvZA8l",
"N6A451n51Ed",
"lNwj0Lrufb",
"lNwj0Lrufb",
"lNwj0Lrufb",
"uSSf9VZAo-k",
"BSAJ6QD7gaS",
"nips_2022_Vhg... |
nips_2022_rWgfLdqVVl_ | Visual Concepts Tokenization | Obtaining the human-like perception ability of abstracting visual concepts from concrete pixels has always been a fundamental and important target in machine learning research fields such as disentangled representation learning and scene decomposition. Towards this goal, we propose an unsupervised transformer-based Vis... | Accept | This paper proposes an unsupervised transformer based framework called Visual Concepts Tokenization (VCT) to extract visual concepts from concrete pixels for tackling disentangled representation learning and scene decomposition. Experiments on several popular datasets validated the effectiveness of VCT on the tasks of ... | train | [
"bc56DbjYPKo",
"jy3FH6FW6vX",
"VtZ2Ppoiolp",
"8a3YI6vyzST",
"-_GNt9dNbXw",
"OIoXPsCcBa",
"axKfB2z49R",
"5IhwGBwnaRh",
"LL_EBAusiCb",
"lISA7-OS_Xr",
"qveEFlrv1L",
"aSomoh-iMUx",
"0z9uBf76IP-",
"ktp95JErHSc",
"ActI4mOS43t",
"k3-VjkAMY-0",
"Te6zzn1HE6",
"W7YQqOgf7Ab",
"I4H-S6LwU1R"
... | [
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear Reviewer yDux,\n\nWe want to send you a kindly reminder for the discussion, since the stage of discussion will be soon concluded.\n\nWe thank you again for your valuable comments, and we are happy to extend our response if you have any other concerns left.\n\nThanks.",
" Dear Reviewer YRhc,\n\nWe want to s... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
7,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
3
] | [
"qveEFlrv1L",
"axKfB2z49R",
"OIoXPsCcBa",
"axKfB2z49R",
"OIoXPsCcBa",
"ktp95JErHSc",
"ActI4mOS43t",
"qveEFlrv1L",
"W7YQqOgf7Ab",
"Te6zzn1HE6",
"0z9uBf76IP-",
"Te6zzn1HE6",
"I4H-S6LwU1R",
"W7YQqOgf7Ab",
"Te6zzn1HE6",
"nips_2022_rWgfLdqVVl_",
"nips_2022_rWgfLdqVVl_",
"nips_2022_rWgfL... |
nips_2022_GGtH47T31ZC | Orthogonal Transformer: An Efficient Vision Transformer Backbone with Token Orthogonalization | We present a general vision transformer backbone, called as Orthogonal Transformer, in pursuit of both efficiency and effectiveness. A major challenge for vision transformer is that self-attention, as the key element in capturing long-range dependency, is very computationally expensive for dense prediction tasks (e.g.,... | Accept | The paper presents orthogonal attention mechanism for vision transformers. All reviewers found the overall system has good performance and the introduced orthogonal attention has the potential to be widely used. The authors' rebuttal resolves the majority of the questions.
The authors should add their promised additio... | train | [
"-DQoS1bMu3",
"rq0E2qNW1Cl",
"FrAHCTU86Cb",
"Vpipui5yqbw",
"xebHUytYJim",
"4EtRy1xBpnY",
"nBLDZOpLEK",
"CDjIURn-UzT",
"XvY-nWgGhUQ",
"OpeZGExH5jL"
] | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We sincerely thank the Area Chair and the reviewers for your efforts and valuable comments. We thank you for recognizing the positive aspects of our paper, such as the novelty of the orthogonal attention, the extensiveness of experiments, and the impressive performance of our models. \n\nThe major concern is that... | [
-1,
-1,
-1,
-1,
-1,
-1,
4,
5,
5,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
4,
5,
4,
4
] | [
"rq0E2qNW1Cl",
"nips_2022_GGtH47T31ZC",
"OpeZGExH5jL",
"XvY-nWgGhUQ",
"CDjIURn-UzT",
"nBLDZOpLEK",
"nips_2022_GGtH47T31ZC",
"nips_2022_GGtH47T31ZC",
"nips_2022_GGtH47T31ZC",
"nips_2022_GGtH47T31ZC"
] |
nips_2022_Aisi2oEq1sc | Finding Differences Between Transformers and ConvNets Using Counterfactual Simulation Testing | Modern deep neural networks tend to be evaluated on static test sets. One shortcoming of this is the fact that these deep neural networks cannot be easily evaluated for robustness issues with respect to specific scene variations. For example, it is hard to study the robustness of these networks to variations of object ... | Accept | After the rebuttal and discussion all reviewers are positive, and recommend acceptance. The AC agrees with this recommendation. | val | [
"t2r3KdfkACi",
"aio625Nr5J0",
"KrX1dIF8kOj",
"qWRkqBKlcw",
"meOucUMY8hw",
"cQ-1ZZSOFx",
"VKckjHSv7TL",
"Z60R7HNHSH",
"6F4vJOC4Kjc",
"1zQQSIMfVuH",
"BPCHAZC_hdM",
"WlPvjYK4bcM",
"7q_fREyNQs3",
"LiMxGPuSgw"
] | [
"official_reviewer",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" I appreciate the authors' response to my questions and comments and their speed in updating their submission. I think this is a well-written study on differences between ViTs and CNNs. I currently feel though that without an experiment that grounds these results on real-world data, I have some uncertainty about h... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
5,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4
] | [
"BPCHAZC_hdM",
"KrX1dIF8kOj",
"qWRkqBKlcw",
"meOucUMY8hw",
"cQ-1ZZSOFx",
"LiMxGPuSgw",
"7q_fREyNQs3",
"7q_fREyNQs3",
"7q_fREyNQs3",
"WlPvjYK4bcM",
"WlPvjYK4bcM",
"nips_2022_Aisi2oEq1sc",
"nips_2022_Aisi2oEq1sc",
"nips_2022_Aisi2oEq1sc"
] |
nips_2022_juE5ErmZB61 | Polynomial Neural Fields for Subband Decomposition and Manipulation | Neural fields have emerged as a new paradigm for representing signals, thanks to their ability to do it compactly while being easy to optimize. In most applications, however, neural fields are treated like a black box, which precludes many signal manipulation tasks. In this paper, we propose a new class of neural field... | Accept | There is a clear consensus for accepting the paper. The area chair agrees with the reviewer's comments and follows their recommendation. | train | [
"9lZTMfwB_94",
"bCdTmPQEfom",
"kMU9WB6vdgC",
"qvfqQdvox6",
"s0ahE_PGLub",
"QIjksSHMGn",
"GEfT_QIv5N3",
"SBG4MLQc9gd",
"tmHhF0_Dwj"
] | [
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We thank the reviewer for the overall positive and constructive feedback. \n\n**Efficiency analysis**\nPlease refer to our general comment for a detailed discussion of parameter and computation efficiency analysis. In brief, PNF achieves comparable performance with SOTA methods using a similar number of parameter... | [
-1,
-1,
-1,
-1,
-1,
7,
5,
8,
7
] | [
-1,
-1,
-1,
-1,
-1,
3,
3,
3,
3
] | [
"GEfT_QIv5N3",
"tmHhF0_Dwj",
"QIjksSHMGn",
"SBG4MLQc9gd",
"nips_2022_juE5ErmZB61",
"nips_2022_juE5ErmZB61",
"nips_2022_juE5ErmZB61",
"nips_2022_juE5ErmZB61",
"nips_2022_juE5ErmZB61"
] |
nips_2022_80RnitDehg_ | Anticipating Performativity by Predicting from Predictions | Predictions about people, such as their expected educational achievement or their credit risk, can be performative and shape the outcome that they are designed to predict. Understanding the causal effect of predictions on the eventual outcomes is crucial for foreseeing the implications of future predictive models and ... | Accept | Strengths:
* problem is well motivated and stated, writing is clear
* interesting and important identification problem
* useful results on sufficient conditions
* causal assumptions are made explicit (see also weaknesses)
Weaknesses:
* very strict assumptions on causal structure (specifically that $\hat{Y}$ does not a... | train | [
"EgHol-IbRp",
"fbFPZPbTX-z",
"GGFsuVhoV8k",
"MI3Mrfetp5",
"rkAWJyIRD5b",
"fMtE2nsu6O",
"2r6XCRmCrdL",
"M3HWdw_ofkM",
"bThVGo4IEJl",
"J8RYKCmnFqK",
"_vPlgrj9kI",
"mr3potzWkBk",
"jjh2upkWZGw",
"lbksmBhdv_b",
"M6EKD9caBcn",
"Tl9dRd_sKc-",
"-FMBHZ9ujnM",
"QDE3XCY-_A",
"iCdhSP9qf9T"
] | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_re... | [
" We appreciate your positive assessment and the helpful comments. We will for sure implement all these changes in the camera-ready version to the best of our abilities. \n\nRegarding your last question: \nthe fact that $U$ is unobserved means that, for each individual, we do not observe their precise value of $U$,... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
4,
6,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4,
4
] | [
"MI3Mrfetp5",
"rkAWJyIRD5b",
"fMtE2nsu6O",
"J8RYKCmnFqK",
"mr3potzWkBk",
"M3HWdw_ofkM",
"mr3potzWkBk",
"_vPlgrj9kI",
"jjh2upkWZGw",
"iCdhSP9qf9T",
"QDE3XCY-_A",
"-FMBHZ9ujnM",
"Tl9dRd_sKc-",
"M6EKD9caBcn",
"nips_2022_80RnitDehg_",
"nips_2022_80RnitDehg_",
"nips_2022_80RnitDehg_",
"... |
nips_2022_acKK8MQe2xc | Learning Invariant Graph Representations for Out-of-Distribution Generalization | Graph representation learning has shown effectiveness when testing and training graph data come from the same distribution, but most existing approaches fail to generalize under distribution shifts. Invariant learning, backed by the invariance principle from causality, can achieve guaranteed generalization under distri... | Accept | This paper focuses on a new research problem of learning invariant graph representations under distribution shifts, which considers the latent environment labels. The proposal is a joint learning framework called graph invariant learning (GIL), combing three different GNNs of various functions. The philosophy behind so... | train | [
"pUCMcPpMK2",
"YGodvJhIrvJ",
"PM2cQMf8HKZ",
"ARzdGQHR9hm",
"Mmwf99OCHD",
"CdUmIRbJyV",
"Y5HQPkq7EqV",
"gOagjfmarRc",
"Rg03B6e8LUZ",
"xtuXSXIqWlv",
"GaZr8YurMY7",
"RznFYbDCwiF",
"ARsuAsGzlXM",
"GOfSbj-TIyh",
"N15tmanQ9_b",
"cqc-PU8UmAP",
"koKszwiuI5",
"J0CddhRKskH",
"yRJNx8ygJ4w",... | [
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"officia... | [
" Dear Reviewer,\n\nWe sincerely appreciate your positive feedback and acknowledgement of our rebuttal. And thanks again for your time and efforts in reviewing our work.\n\nBest regards,\n\nThe Authors",
" Thank you for the detailed responses to all points made in the review. \n\nWe have differing opinions on the... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
5,
6,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
3,
4,
3
] | [
"YGodvJhIrvJ",
"848hHcGECpW",
"ARzdGQHR9hm",
"S4ICvql11X6",
"S4ICvql11X6",
"848hHcGECpW",
"AtIJ3WLKRzU",
"9RUXsIGR37",
"nips_2022_acKK8MQe2xc",
"GaZr8YurMY7",
"RznFYbDCwiF",
"ARsuAsGzlXM",
"9RUXsIGR37",
"N15tmanQ9_b",
"cqc-PU8UmAP",
"koKszwiuI5",
"848hHcGECpW",
"AtIJ3WLKRzU",
"Z4... |
nips_2022_VgOw1pUPh97 | SegNeXt: Rethinking Convolutional Attention Design for Semantic Segmentation | We present SegNeXt, a simple convolutional network architecture for semantic segmentation. Recent transformer-based models have dominated the field of se- mantic segmentation due to the efficiency of self-attention in encoding spatial information. In this paper, we show that convolutional attention is a more efficient ... | Accept | Four knowledgeable referees reviewed this submission. The reviews raised concerns about the novelty of the proposed approach (rY1T, S4w5), the motivation of the model design and properties (mij9, r2Bq), and the empirical evidence to support some of the effectiveness and efficiency claims (rY1T, r2Bq, S4w5). The rebutta... | train | [
"Jah_yH_Ep8V",
"HzXXm-JpQg",
"aAiqJ6W9pGD",
"dSjLhFb2lcA",
"hHebw5_hwR",
"7alClLbYRpd",
"1_X16sO8p3A",
"Hh7sdxOekoB",
"IQfBgAE7AEr0",
"8GnrZubWs85",
"0gRf9j85oje",
"bc4_lyvRrjr",
"-bw-925kVgV",
"i-kDMUUMvDI",
"TaROyTlWYj",
"nHFfaH-IhX6",
"8Aqlk1BEqCk",
"TxMqvoogAjt",
"__Q5h7420zi... | [
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" In fact, we had organized the related code like SegFormer[1], which can be released at any time.\n\nDue to the anonymity rules of NeurIPs 2022, we can not provide it now.\n\nTrust us, the code will be public as mentioned in abstract. \n\n[1]: Xie, E., Wang, W., Yu, Z., Anandkumar, A., Alvarez, J. M., & Luo, P. (2... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
7,
5,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
4,
5
] | [
"aAiqJ6W9pGD",
"aAiqJ6W9pGD",
"i-kDMUUMvDI",
"hHebw5_hwR",
"nHFfaH-IhX6",
"1_X16sO8p3A",
"bc4_lyvRrjr",
"IQfBgAE7AEr0",
"8GnrZubWs85",
"0gRf9j85oje",
"__Q5h7420zi",
"TxMqvoogAjt",
"8Aqlk1BEqCk",
"TaROyTlWYj",
"nHFfaH-IhX6",
"nips_2022_VgOw1pUPh97",
"nips_2022_VgOw1pUPh97",
"nips_20... |
nips_2022_Iqm6AiHPs_z | Active Labeling: Streaming Stochastic Gradients | The workhorse of machine learning is stochastic gradient descent.
To access stochastic gradients, it is common to consider iteratively input/output pairs of a training dataset.
Interestingly, it appears that one does not need full supervision to access stochastic gradients, which is the main motivation of this paper.
A... | Accept | This paper studies “active labeling”, which can be seen as active learning with weak supervision, and proposes an active labeling algorithm based on SGD. The reviewers found that the idea of this paper is innovative. After author response and reviewer discussion, the paper receives generally unanimous support from the ... | val | [
"NWoMrpbyBP1",
"FedXKjQM_fJ",
"nxuvBw3JqD",
"7jwYjcs81H",
"XK4H1LW998d",
"SFPzDkNMZFx",
"lItw4QUXmR3K",
"6OYDHlwaCA-",
"sgPwtBGF6L5",
"rofJrX8bm2K",
"07kDQVmWBS4",
"i5dt1_rcr7wX",
"5DKHTxflsyt7",
"k1GIGXfH-L",
"Pqf8oY1mEaY",
"WkD60q7HR35",
"3aSiRauBid"
] | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you again for your productive comments, we highly appreciate your inputs.\nBe sure that we will do our best to integrate them in order to help future readers to get the most of our work.\nIn particular, we will make section 7 more precise, discuss the ranking example, and give more motivations for the vecto... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
5,
6,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
2,
4,
4,
3
] | [
"SFPzDkNMZFx",
"nxuvBw3JqD",
"XK4H1LW998d",
"5DKHTxflsyt7",
"i5dt1_rcr7wX",
"6OYDHlwaCA-",
"rofJrX8bm2K",
"sgPwtBGF6L5",
"07kDQVmWBS4",
"k1GIGXfH-L",
"Pqf8oY1mEaY",
"WkD60q7HR35",
"3aSiRauBid",
"nips_2022_Iqm6AiHPs_z",
"nips_2022_Iqm6AiHPs_z",
"nips_2022_Iqm6AiHPs_z",
"nips_2022_Iqm6... |
nips_2022_lTKXh991Ayv | Practical Adversarial Attacks on Spatiotemporal Traffic Forecasting Models | Machine learning based traffic forecasting models leverage sophisticated spatiotemporal auto-correlations to provide accurate predictions of city-wide traffic states. However, existing methods assume a reliable and unbiased forecasting environment, which is not always available in the wild. In this work, we investigate... | Accept | The authors have made a significant effort to address reviewer concerns in their rebuttal. They are strongly encouraged to include these additional results and observations to either the main body of the paper or the supplement. | train | [
"M-MxVjuosoD",
"_lSpVkYHyDd",
"u9GhkcyuUY",
"GxFdze3ideb",
"FrJn64E5k5u",
"L9vcLSPSDV3",
"L_QEX6OyCxu",
"aX2Q9Fnc2WQ",
"8JED3JT3ga1",
"wlI11gp_bv9",
"xUMTdnYaFTD",
"vL_4Mf6QpgA",
"pCsdlz5zQG",
"ApKtVd6oTyV",
"fdEpIPjCWaY",
"YctMpq3Nzgf"
] | [
"author",
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear reviewer rQBx:\n\nWe thank you for the precious review time and valuable comments. We have provided corresponding responses and results, which we believe have covered your concerns. We hope to further discuss with you whether or not your concerns have been addressed. Please let us know if you still have any ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
4,
6,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
3,
2,
4
] | [
"fdEpIPjCWaY",
"ApKtVd6oTyV",
"pCsdlz5zQG",
"FrJn64E5k5u",
"L9vcLSPSDV3",
"L_QEX6OyCxu",
"YctMpq3Nzgf",
"fdEpIPjCWaY",
"ApKtVd6oTyV",
"ApKtVd6oTyV",
"ApKtVd6oTyV",
"pCsdlz5zQG",
"nips_2022_lTKXh991Ayv",
"nips_2022_lTKXh991Ayv",
"nips_2022_lTKXh991Ayv",
"nips_2022_lTKXh991Ayv"
] |
nips_2022_Pyd6Rh9r1OT | Fast Vision Transformers with HiLo Attention | Vision Transformers (ViTs) have triggered the most recent and significant breakthroughs in computer vision. Their efficient designs are mostly guided by the indirect metric of computational complexity, i.e., FLOPs, which however has a clear gap with the direct metric such as throughput. Thus, we propose to use the dire... | Accept | Initially, this paper received diverging reviews. The authors did a good job addressing the reviewers' concerns, by adding additional comparisons to more SOTA ViT backbones and benchmarking throughput on a variety of GPU platforms. The AC agrees with the reviewer that the concerns have been sufficiently addressed and r... | train | [
"xqsZ-aWVZK",
"zlAsVLwaqB4",
"eVKnRobml6E",
"d6nmErtW3nq",
"5Fg8D_Qw81M",
"AdK3j7luoXf",
"gzNUNuC10cK",
"hQ01F0k3wu-",
"893K1zwx0ds",
"oDdvL-jUdWE",
"qAsfQuasBpi",
"sUA9DglLox-",
"b-NSXlEZCBi",
"wZipyx_eCfk",
"HJG86cW29Nj",
"ceDxNLbuWa3"
] | [
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for your valuable feedback and suggestions! We are glad to address your questions and appreciate your constructive reviews for improving our work. ",
" Dear authors,\n\nThank you for your comments, added experiments (table IV) and explanations. You have addressed my main concerns as I wrote in detail in ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
6,
9,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
5,
5
] | [
"zlAsVLwaqB4",
"oDdvL-jUdWE",
"d6nmErtW3nq",
"893K1zwx0ds",
"ceDxNLbuWa3",
"wZipyx_eCfk",
"ceDxNLbuWa3",
"ceDxNLbuWa3",
"HJG86cW29Nj",
"wZipyx_eCfk",
"b-NSXlEZCBi",
"b-NSXlEZCBi",
"nips_2022_Pyd6Rh9r1OT",
"nips_2022_Pyd6Rh9r1OT",
"nips_2022_Pyd6Rh9r1OT",
"nips_2022_Pyd6Rh9r1OT"
] |
nips_2022_rA2tItoRUth | LGDN: Language-Guided Denoising Network for Video-Language Modeling | Video-language modeling has attracted much attention with the rapid growth of web videos. Most existing methods assume that the video frames and text description are semantically correlated, and focus on video-language modeling at video level. However, this hypothesis often fails for two reasons: (1) With the rich sema... | Accept | This paper proposes Language-Guided Denoising Network (LGDN) with the goal of addressing the redundancy and noisy alignment issues in video-language modeling, by dynamically filtering out misaligned or redundant frames under the language supervision and obtains only 2-4 salient frames per video for cross-modal token-le... | train | [
"17IPPG0Yp09",
"Gzn5HqHDjK4",
"RTqRCtwsvCI",
"3eAAsEcMk_",
"PaEG3FkewT",
"4zwRQcSJo5d",
"S8dhD7BFubk",
"leRUHlfatRi",
"jOvjAJSk7y2",
"VZGEnni_fpqC",
"BkDwzSybzPfP",
"ea59EvfnAby",
"w2ncheR8BHM",
"xGrXlYFsG7A",
"d9Q10_TUdMR",
"pCgvuv_uFgJ",
"c6WO4PtpRYv",
"CfmlGTnpaDU"
] | [
"official_reviewer",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" I would like to thank the authors for addressing all of my comments. And, the added limitations and potential negative societal impacts are reasonable to me. Thank you!",
" Dear Reviewer HH2K:\n\nThanks again for spending a huge amount of time on our paper, which has helped us improve the quality and clarity of... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
6,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
4
] | [
"d9Q10_TUdMR",
"RTqRCtwsvCI",
"3eAAsEcMk_",
"PaEG3FkewT",
"S8dhD7BFubk",
"CfmlGTnpaDU",
"leRUHlfatRi",
"jOvjAJSk7y2",
"c6WO4PtpRYv",
"nips_2022_rA2tItoRUth",
"nips_2022_rA2tItoRUth",
"CfmlGTnpaDU",
"c6WO4PtpRYv",
"c6WO4PtpRYv",
"pCgvuv_uFgJ",
"nips_2022_rA2tItoRUth",
"nips_2022_rA2tI... |
nips_2022_7YTh6S8HIY | PyramidCLIP: Hierarchical Feature Alignment for Vision-language Model Pretraining | Large-scale vision-language pre-training has achieved promising results on downstream tasks. Existing methods highly rely on the assumption that the image-text pairs crawled from the Internet are in perfect one-to-one correspondence. However, in real scenarios, this assumption can be difficult to hold: the text descrip... | Accept | This paper proposes PyramidCLIP. It improve contrastive learning method CLIP with more fine-grained information to produce multiple views of both the image and text during training. During inference/evaluation, only the standard view is used. The empirical results with different network architectures at different pret... | train | [
"_HxXP2hX8m",
"qK-T1AhgHB",
"eTzZUaOe6ST",
"lnKxcTG1058",
"haZ5FrWdSrG",
"k7nj5tX9GBu",
"FZm2TPvUyB0",
"6uxKcYaFXQy",
"VEj5bsfnO6e",
"229MPfXBWE6",
"yc7NdZbq2uv",
"HQg4dqN9DD",
"1b2Sx1dcIF",
"4iG29vcdk3Y",
"S-KKhEPq7g8",
"npDczYLTMFd",
"QIC4i5R3sUc",
"Ch9vFhjwDMg",
"Ib-op6cQ7lM",... | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
... | [
" Thank you for the response!\n\nThe qualitative examples do address my concerns partially. I am still a bit uncomfortable with using summarization models that are pre-trained for summarizing long documents (>400 words) in the news domain in summarizing short captions (10~30 words) in another domain, especially con... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
5,
8,
5,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
4,
3
] | [
"haZ5FrWdSrG",
"lnKxcTG1058",
"QIC4i5R3sUc",
"FZm2TPvUyB0",
"k7nj5tX9GBu",
"HQg4dqN9DD",
"VEj5bsfnO6e",
"Ib-op6cQ7lM",
"K2IfD7o8HY",
"K2IfD7o8HY",
"Ib-op6cQ7lM",
"Ib-op6cQ7lM",
"Ch9vFhjwDMg",
"QIC4i5R3sUc",
"npDczYLTMFd",
"nips_2022_7YTh6S8HIY",
"nips_2022_7YTh6S8HIY",
"nips_2022_7... |
nips_2022_NjImFaBEHl | Divide and Contrast: Source-free Domain Adaptation via Adaptive Contrastive Learning | We investigate a practical domain adaptation task, called source-free domain adaptation (SFUDA), where the source pretrained model is adapted to the target domain without access to the source data. Existing techniques mainly leverage self-supervised pseudo-labeling to achieve class-wise global alignment [1] or rely on ... | Accept | This paper proposes a relatively complicated method for source-free unsupervised domain adaptation, which integrates several techniques into a divide and contrast framework. The idea of dividing the target data into source-like subset and target-specific subset and employing global alignment and feature consistency for... | val | [
"fxIUL2OxPjM",
"uLx_EzivY-Q",
"mHxVheo7uG",
"aMi-dDPorsh",
"ovB5_dNXmA8",
"t7HHbk5V-NN",
"EB8QgLya_H8",
"b73UJPbAfGX",
"lPffnudFu1g",
"Cp36eEwy2HX",
"Fq9YKma9Hgt"
] | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Authors have addressed the error bars issue that I have raised. Other reviewers have raised some concerns about comparisons with SOTA approaches, but it appears that the authors have adequately addressed most of the concerns.",
" Thanks for the reply. My concerns are addressed.",
" We thank reviewer xRX9 for ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
5,
5,
8
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
5,
3
] | [
"EB8QgLya_H8",
"aMi-dDPorsh",
"b73UJPbAfGX",
"b73UJPbAfGX",
"lPffnudFu1g",
"Cp36eEwy2HX",
"Fq9YKma9Hgt",
"nips_2022_NjImFaBEHl",
"nips_2022_NjImFaBEHl",
"nips_2022_NjImFaBEHl",
"nips_2022_NjImFaBEHl"
] |
nips_2022_S7Evzt9uit3 | Mind the Gap: Understanding the Modality Gap in Multi-modal Contrastive Representation Learning | We present modality gap, an intriguing geometric phenomenon of the representation space of multi-modal models. Specifically, we show that different data modalities (e.g. images and text) are embedded at arm's length in their shared representation in multi-modal models such as CLIP. Our systematic analysis demonstrates ... | Accept | This paper investigates the gap between representations when training with a contrastive objective, through the characterisation of the gap in various settings, and building a theoretical analysis of this gap.
The reviewers mostly agree that the paper tackles an interesting problem through investigation and characteri... | train | [
"aSO1uIdCtwN",
"n3pxQC95RQL",
"7TeXh8t-4yN",
"rrrdrDfWDtT",
"EQSHdKkACf9",
"NmW0pPUeEUj9",
"y4rxL0il36G",
"0wJY-Sg6her",
"xeNHouujcM9",
"ZvNm1vLb5S0",
"vkZfDby3uwi",
"yKhOe6RPl8C",
"nrMw71JSD1",
"33eEUPW5oob",
"1pDmowld46R",
"OtLjxYKggP2",
"KhmazHX8O-N",
"8uVCttIGer",
"i-Ozc_cWG0... | [
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear reviewer QxpT,\n\nWe would like to follow up to see if our response addresses your concerns or if you have any further questions. We would really appreciate the opportunity to discuss this further if our response has not already addressed your concerns. Thank you very much!",
" Dear reviewer nReN,\n\nWe wo... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
6,
6,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
3,
5
] | [
"5NQDLW6Q9r3",
"i-Ozc_cWG0R",
"8uVCttIGer",
"EQSHdKkACf9",
"NmW0pPUeEUj9",
"y4rxL0il36G",
"KhmazHX8O-N",
"xeNHouujcM9",
"nips_2022_S7Evzt9uit3",
"vkZfDby3uwi",
"i-Ozc_cWG0R",
"nrMw71JSD1",
"8uVCttIGer",
"1pDmowld46R",
"KhmazHX8O-N",
"5NQDLW6Q9r3",
"nips_2022_S7Evzt9uit3",
"nips_202... |
nips_2022_OjS3nkNATOw | Adapting Self-Supervised Vision Transformers by Probing Attention-Conditioned Masking Consistency | Visual domain adaptation (DA) seeks to transfer trained models to unseen, unlabeled domains across distribution shift, but approaches typically focus on adapting convolutional neural network architectures initialized with supervised ImageNet representations. In this work, we shift focus to adapting modern architectures... | Accept | This work looks at adapting ViT-like models for unsupervised domain adaptation, by cleverly finding pseudo-labels with 'attention-guided masking'. There's a weak consensus among the reviewers that this work has good empirical results, but somewhat limited novelty. I think the rebuttal discussion has helped improve this... | train | [
"pulj6WK56mN",
"HoKoEfVLneA",
"vT1x6wAnXC",
"0oId3uOtpCJ",
"3jfCTIibVfs",
"xGaks6sDs7i",
"k5byWRtIXBS",
"sA178xhXoe8",
"3zCqOr_fwhu",
"N6x-1ZE7NfS",
"IKyp9sc4zcL",
"cPNDJioZWKx",
"sCClK7Rz4CN"
] | [
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We would be happy to address any additional questions. Otherwise, we would appreciate if the reviewer would consider updating their score in light of the clarifications and new experiments.\n\n",
" We would be happy to address any additional questions. Otherwise, we'd appreciate if the reviewer would consider u... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
6,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4
] | [
"3zCqOr_fwhu",
"0oId3uOtpCJ",
"3jfCTIibVfs",
"xGaks6sDs7i",
"k5byWRtIXBS",
"sA178xhXoe8",
"sCClK7Rz4CN",
"cPNDJioZWKx",
"IKyp9sc4zcL",
"nips_2022_OjS3nkNATOw",
"nips_2022_OjS3nkNATOw",
"nips_2022_OjS3nkNATOw",
"nips_2022_OjS3nkNATOw"
] |
nips_2022_jowVZoitZYu | On Trace of PGD-Like Adversarial Attacks | Adversarial attacks pose safety and security concerns for deep learning applications. Yet largely imperceptible, a strong PGD-like attack may leave strong trace in the adversarial example. Since attack triggers the local linearity of a network, we speculate network behaves in different extents of linearity for benign... | Reject | This paper observes that "PGD-like" attack algorithms have characteristics that
allow one to detect an input has been attacked. While I agree that it is
interesting PGD-like attacks have detectable properties, I agree with the
reviewers that designing attck-specific defenses has limited utility. The
authors already sho... | train | [
"vdM6PKmF7a3",
"0JXUU_EXXWRM",
"j-z8-gwy5df",
"bzdAzYjHiXg",
"Zpo4AlOtcuF",
"cAY1s_jUHgM",
"vdL0qcqgABX",
"hq4-fNXeXfZ",
"4nM9clN-Cw4",
"KO81F-sNIpD",
"J5pSDsIDhTC",
"3IMQZhseHd7",
"dkCYvOsFR8J",
"4OKs4YihynZ",
"ocrOggOlRQX",
"ks3TllR9f2U",
"zLbl2g-jOPx",
"N6w3yvLDbBY",
"J_Khri3h... | [
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_re... | [
" Thank the authors for the detailed response and comprehensive discussions. The proposed method is relatively new and interesting, which could potentially arouse future research along this line. Meanwhile, the application of the current method is somewhat limited as pointed out by the other reviewers. Therefore, I... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
6,
6,
4
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
2,
4,
5,
4
] | [
"4OKs4YihynZ",
"bzdAzYjHiXg",
"KO81F-sNIpD",
"4nM9clN-Cw4",
"vdL0qcqgABX",
"vdL0qcqgABX",
"3IMQZhseHd7",
"J5pSDsIDhTC",
"KO81F-sNIpD",
"ocrOggOlRQX",
"dkCYvOsFR8J",
"N6w3yvLDbBY",
"J_Khri3h609",
"zLbl2g-jOPx",
"ks3TllR9f2U",
"nips_2022_jowVZoitZYu",
"nips_2022_jowVZoitZYu",
"nips_2... |
nips_2022_m67FNFdgLO9 | Dense Interspecies Face Embedding | Dense Interspecies Face Embedding (DIFE) is a new direction for understanding faces of various animals by extracting common features among animal faces including human face. There are three main obstacles for interspecies face understanding: (1) lack of animal data compared to human, (2) ambiguous connection between fa... | Accept | This paper uses knowledge distillation to transfer learn facial embeddings across humans and animals. Helps when sufficient data for learning embeddings from animal faces is not available. An interesting application of standard concepts from domain adaptation, knowledge distillation, etc. While preparing the final pape... | train | [
"LW04V-f_nvi",
"VGQlcNLR1hM",
"aYXt8CfmH1J",
"iFaMW8IArv6",
"hrj6bfislK",
"yCh0N5kBLv",
"1-BayH1bvoU",
"ZtGEzoo0Nm9",
"PalDNHeERqa",
"nyMveVtmfjJ",
"n3mlSG42241",
"VQstlncmXJv",
"0K8kgcfEz9i"
] | [
"official_reviewer",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for your answers. They addressed most of my concerns pre-rebuttal. I will update my score.\n\n",
" ### Table 3: The human landmark detection\n- We bring citations starting with ‘O’ and the NME value of previous works from On Equivariant and Invariant [R1]. For example, ‘O67’ means the 67th reference of t... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
6,
6,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
3
] | [
"ZtGEzoo0Nm9",
"iFaMW8IArv6",
"nips_2022_m67FNFdgLO9",
"yCh0N5kBLv",
"0K8kgcfEz9i",
"VQstlncmXJv",
"n3mlSG42241",
"nyMveVtmfjJ",
"nips_2022_m67FNFdgLO9",
"nips_2022_m67FNFdgLO9",
"nips_2022_m67FNFdgLO9",
"nips_2022_m67FNFdgLO9",
"nips_2022_m67FNFdgLO9"
] |
nips_2022_H3JObxjd8S | Self-Supervised Visual Representation Learning with Semantic Grouping | In this paper, we tackle the problem of learning visual representations from unlabeled scene-centric data. Existing works have demonstrated the potential of utilizing the underlying complex structure within scene-centric data; still, they commonly rely on hand-crafted objectness priors or specialized pretext tasks to b... | Accept | This paper proposes an object-centric representation learning based on a data-driven semantic slots from scene-centric data. In specific, the proposed SlotCon simultnesobly performs semantic grouping and contrastive representation learning over groups (slots), which naturally leads to obtaining object-level representat... | train | [
"xe3S7S4E8fe",
"xBl5UdKFLdw",
"qiFH23XpXxW",
"hnQHxUnLLE_",
"Z7ZOMLoPjPl",
"WEYyPBhKQkq",
"wxjpN5wHYgZ",
"87dwEGTP75o",
"KTb9huf6V0",
"RcFqzT3FFUG",
"sDnGhRDEnS",
"5IAG0HcYYMjD",
"Bg0xtGqcMFfH",
"q8cxHaptNfd",
"Y6-KBBHctDu",
"aaPQLtPU-Du",
"L6PBSjATGb1",
"6o6MoJFFgRmz",
"ONPwqLsM... | [
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_... | [
" Dear Reviewer NYjs,\n\nThank you for the time you devoted to reviewing this paper and your further constructive comments in the prior issue.\n\nAfter carefully reading your comments and reflecting on the experiment results, we agree that our previous statements regarding the priors are too strong and the scope is... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
5,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
3,
4
] | [
"hnQHxUnLLE_",
"Z7ZOMLoPjPl",
"nips_2022_H3JObxjd8S",
"ONPwqLsMIZM",
"Y6-KBBHctDu",
"KTb9huf6V0",
"RcFqzT3FFUG",
"sDnGhRDEnS",
"q8cxHaptNfd",
"Bg0xtGqcMFfH",
"6o6MoJFFgRmz",
"nips_2022_H3JObxjd8S",
"KJn_YdmDjoO",
"KJn_YdmDjoO",
"CltN9Irkkz",
"_mLQ_283L-w",
"_mLQ_283L-w",
"_mLQ_283L... |
nips_2022_uV_VYGB3FCi | Flexible Neural Image Compression via Code Editing | Neural image compression (NIC) has outperformed traditional image codecs in rate-distortion (R-D) performance. However, it usually requires a dedicated encoder-decoder pair for each point on R-D curve, which greatly hinders its practical deployment. While some recent works have enabled bitrate control via conditional c... | Accept | Thanks for your submission to NeurIPS. The reviewers are all in agreement that the paper is ready for publication. They in particular appreciated your rebuttals and changes to the paper, and increased their scores as a result. The proposed method is novel, interesting, and performs well. | train | [
"5xf2LelLZEC",
"zvoHPN_KTN",
"uPUOFp4mrZU",
"yeklw7DVh_Y",
"0TWWsPGBX3C",
"puvrT_n-kle",
"JkP74r1qiE",
"mn-Fn6D-GzcB",
"cNGrLqwt-fV",
"l4m8-spFs1z",
"l9VGqdAoNT7"
] | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you very much for your kind additional explanation. Your additional explanation and revision helped me understand your claim (Yang et al. 2020 requires multiple models, whereas the proposed method can achieve nearly the same RD-rate with a single model). I will raise my evaluation of the paper.",
" Thanks... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
8
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
5
] | [
"0TWWsPGBX3C",
"mn-Fn6D-GzcB",
"nips_2022_uV_VYGB3FCi",
"l9VGqdAoNT7",
"cNGrLqwt-fV",
"cNGrLqwt-fV",
"l4m8-spFs1z",
"l4m8-spFs1z",
"nips_2022_uV_VYGB3FCi",
"nips_2022_uV_VYGB3FCi",
"nips_2022_uV_VYGB3FCi"
] |
nips_2022_8rZYMpFUgK | DAGMA: Learning DAGs via M-matrices and a Log-Determinant Acyclicity Characterization | The combinatorial problem of learning directed acyclic graphs (DAGs) from data was recently framed as a purely continuous optimization problem by leveraging a differentiable acyclicity characterization of DAGs based on the trace of a matrix exponential function. Existing acyclicity characterizations are based on the id... | Accept | Overall, reviews for this paper are quite positive. The paper presents an interesting and effective new approach to incorporating a DAG constraint into an optimization problem by using a characterization of DAGs in terms of the logdet function.
During discussion, the reviewers raised several important questions/points... | test | [
"5DHRulmpAz2",
"v59xslewGw",
"S1afBmjmO2v",
"ZldpXZxiQQ",
"gkZ6n-zJTf4",
"Tu8BNFJpoC",
"x0r1ivAsbhg",
"FVlGg_piuvI",
"1dO_JJ2MK9m",
"sGdluJI7hmF",
"HQ21FsNceOq",
"oFKEgDPCfyXX",
"CweTKY83pBs",
"DVPkMCU0AXX",
"5nd8bVYTszJ",
"-1sAFKRrEvX",
"uhkRSY8eq1",
"9z7JbEr4EDA",
"gJL0fGUUkJG"... | [
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
... | [
" I have updated my overall assessment to \"7: Accept\".",
" We are glad that our response addressed your concerns! And we are highly grateful for your support towards the acceptance of our work!",
" We are happy that our response clarified your doubts! And thank you again for your positive assessment of our wo... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
7,
7,
8
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
3,
3,
3
] | [
"ZldpXZxiQQ",
"Tu8BNFJpoC",
"x0r1ivAsbhg",
"FVlGg_piuvI",
"1dO_JJ2MK9m",
"DVPkMCU0AXX",
"5nd8bVYTszJ",
"CweTKY83pBs",
"sGdluJI7hmF",
"HQ21FsNceOq",
"gJL0fGUUkJG",
"nips_2022_8rZYMpFUgK",
"lggm4rIfvr",
"2eEjhG7YFQ",
"-1sAFKRrEvX",
"PbIfP2QdwUS",
"9z7JbEr4EDA",
"gJL0fGUUkJG",
"T57W... |
nips_2022_xubxAVbOsw | The Minority Matters: A Diversity-Promoting Collaborative Metric Learning Algorithm | Collaborative Metric Learning (CML) has recently emerged as a popular method in recommendation systems (RS), closing the gap between metric learning and Collaborative Filtering. Following the convention of RS, existing methods exploit unique user representation in their model design. This paper focuses on a challenging... | Accept | After the author response, all three reviewers are in favor of accepting the paper. The proposed collaborative metric learning approach was appreciated in terms of novelty by two of the reviewers, whereas one considered the contribution limited. The theoretical analysis was appreciated and the experiment was considered... | train | [
"-uwg0GYytfW",
"3jFpg4zdTJy",
"AVfgQi3jAVU",
"rzER2BXMUZ",
"MVp5z6KOSjJ",
"Mj13W3DY_a",
"_xEICdQCyae",
"ZNYsJMhu_-O",
"Lwj9B9_UL1J",
"8PuRHwDulpH",
"lOd8gVbasjP"
] | [
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" I would like to thank the authors for their thoughtful reply and willingness to meaningfully engage with my questions and concerns. You have gone above and beyond in your response. I believe the additional results provided above significantly strengthen the empirical experiments and further validate the theoretic... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
7,
8
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4
] | [
"Mj13W3DY_a",
"8PuRHwDulpH",
"8PuRHwDulpH",
"8PuRHwDulpH",
"8PuRHwDulpH",
"8PuRHwDulpH",
"lOd8gVbasjP",
"Lwj9B9_UL1J",
"nips_2022_xubxAVbOsw",
"nips_2022_xubxAVbOsw",
"nips_2022_xubxAVbOsw"
] |
nips_2022_CJGUABT_COm | DOMINO: Decomposed Mutual Information Optimization for Generalized Context in Meta-Reinforcement Learning | Adapting to the changes in transition dynamics is essential in robotic applications. By learning a conditional policy with a compact context, context-aware meta-reinforcement learning provides a flexible way to adjust behavior according to dynamics changes. However, in real-world applications, the agent may encounter c... | Accept | This paper proposes DOMINO, an optimization framework, for contextual meta reinforcement learning. The reviewers generally agree that the paper is well written, the idea is novel and interesting, the evaluation is comprehensive and the results are impressive. Reviewers also raised a few concerns in the initial reviews,... | train | [
"znoJkB9IEXT",
"2WhTuSzXj3y",
"A59VmCvb7uD",
"NVvW8P0b2kb",
"5VQTja-eCnV",
"SsKD54e7cMG",
"CxsxixNFzcE",
"ecPVLuuFS",
"sNddMmbI6uo",
"63-Vv8Jz94",
"oWbRzj8sFKW",
"0kZ_C8gf7CB",
"p87kx5ezqV",
"w87doH_JV3t",
"0flbimho4X5",
"Ol0zoOhaM4o",
"J8poTxluEXM",
"llNVuAaJdxf",
"8ltysdzxonvJ"... | [
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
... | [
" We sincerely appreciate all reviewers' and ACs' time and efforts in reviewing our paper. We truly thank you all for the insightful and constructive suggestions, which helped improve the quality of our paper. We appreciate all reviewers have actively discussed with us in-depth and read our responses carefully, whi... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
7,
6,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
5
] | [
"nips_2022_CJGUABT_COm",
"A59VmCvb7uD",
"NVvW8P0b2kb",
"SsKD54e7cMG",
"w87doH_JV3t",
"ecPVLuuFS",
"_bOgSB4vaC",
"0flbimho4X5",
"63-Vv8Jz94",
"p87kx5ezqV",
"J8poTxluEXM",
"Ol0zoOhaM4o",
"Ol0zoOhaM4o",
"rhQabDwjkbv",
"0nq3jEUoxaj",
"rG068puM45P",
"wTJssqx67Bz",
"nips_2022_CJGUABT_COm... |
nips_2022_Ddd6FqHXmHA | OpenAUC: Towards AUC-Oriented Open-Set Recognition | Traditional machine learning follows a close-set assumption that the training and test set share the same label space. While in many practical scenarios, it is inevitable that some test samples belong to unknown classes (open-set). To fix this issue, Open-Set Recognition (OSR), whose goal is to make correct predictio... | Accept | The paper proposes OpenAUC, which is a novel metric designed specifically for evaluating Open-Set Recognition (OSR) performance. OpenAUC is motivated by a formal analysis on existing OSR evaluation metrics, which suffer from three types of inconsistency properties. Theoretical results show that OpenAUC is consistent wi... | train | [
"76y1P2QMX5C",
"IPh0ikjq_-a",
"hZGBKBsaeW0",
"DsePukpCEox",
"fRba-LuY15e",
"O0a-GEfZbin",
"L5jAUPhKxzX",
"IlqeRj0sgez",
"LoeRF1bZ8l",
"_nXcPdXnAF",
"gV7LyA-aXxi",
"Y3TyYjJ8x5o",
"wxegDpgJBbC",
"ew10SzwKNh"
] | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you very much for your nice comments! we would like to make the following responses. \n\n> Comment (a): The open-set function r is the exactly same as \"1 - max-of-softmax-score\", right? If so, it is natural to ask whether other methods such as max-of-logit [14] benefit from the OpenAUC optimized network.\... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
9,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
5,
4
] | [
"DsePukpCEox",
"hZGBKBsaeW0",
"gV7LyA-aXxi",
"L5jAUPhKxzX",
"ew10SzwKNh",
"ew10SzwKNh",
"ew10SzwKNh",
"wxegDpgJBbC",
"Y3TyYjJ8x5o",
"Y3TyYjJ8x5o",
"Y3TyYjJ8x5o",
"nips_2022_Ddd6FqHXmHA",
"nips_2022_Ddd6FqHXmHA",
"nips_2022_Ddd6FqHXmHA"
] |
nips_2022_csr9uRmTC3f | Exploring the Algorithm-Dependent Generalization of AUPRC Optimization with List Stability | Stochastic optimization of the Area Under the Precision-Recall Curve (AUPRC) is a crucial problem for machine learning. Although various algorithms have been extensively studied for AUPRC optimization, the generalization is only guaranteed in the multi-query case. In this work, we present the first trial in the single-... | Accept | All reviewers agree the paper makes novel contributions for AUPRC optimization.
It proposes new batch-based estimator of AUPRC and studies its approximation error. Then it develops a new algorithms for optimizing this estimator. It also establishes the generalization error of the proposed algorithms via novel listwis... | train | [
"TP8_p2EiaQr-",
"fiqbvA-zrr5",
"2zKBGZrfZt8",
"9goRKAqkJzu",
"8bwgJYFnrdb",
"qF_y9WIJ1V1",
"kq61ulbapb6"
] | [
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for your time and positive comments on our manuscript! We would like to reply to the following questions:\n### **Q1:**\nThe mechanism of the semi-variance regulation. Why the improvement in R@K is more significant than mAUPRC?\n### **A1:**\nThe semi-variance regulation is motivated by Prop. 2, which say... | [
-1,
-1,
-1,
-1,
8,
7,
7
] | [
-1,
-1,
-1,
-1,
4,
4,
5
] | [
"8bwgJYFnrdb",
"kq61ulbapb6",
"qF_y9WIJ1V1",
"8bwgJYFnrdb",
"nips_2022_csr9uRmTC3f",
"nips_2022_csr9uRmTC3f",
"nips_2022_csr9uRmTC3f"
] |
nips_2022_MbCAOMGsZXC | Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud Pre-training | Masked Autoencoders (MAE) have shown great potentials in self-supervised pre-training for language and 2D image transformers. However, it still remains an open question on how to exploit masked autoencoding for learning 3D representations of irregular point clouds. In this paper, we propose Point-M2AE, a strong Multi-s... | Accept | This paper proposes, the Point-M2AE, a multi-scale masked autoencoder (MAE) pre-training framework for self-supervised learning of 3D point clouds. This is a generalization of the existing 2D-MAE framework to 3D point cloud domain. The proposed Point-M2AE introduces a U-Net-like transformer and a multi-scale masking st... | train | [
"a87lj0vHEUQ",
"smtfNcXAotK",
"FOf0SPN2bsfs",
"iSIUXmL-zLE",
"zzcL6W9rBIS",
"HXbqeYdA69y",
"UoG_wGDQ23h",
"N28EVxYMt17",
"VjMzVDtyE7p",
"9jxCa8LrMEA"
] | [
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear Reviewers,\n\nThanks again for your insightful comments and valuable time in reviewing our paper. We have provided the corresponding responses to your concerns, and added them in ***the revised supplementary material*** accordingly, which are highlighted in blue.\n\nGiven the discussion phase is quickly pass... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
5,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4
] | [
"nips_2022_MbCAOMGsZXC",
"9jxCa8LrMEA",
"9jxCa8LrMEA",
"VjMzVDtyE7p",
"VjMzVDtyE7p",
"N28EVxYMt17",
"N28EVxYMt17",
"nips_2022_MbCAOMGsZXC",
"nips_2022_MbCAOMGsZXC",
"nips_2022_MbCAOMGsZXC"
] |
nips_2022_u4ihlSG240n | OmniVL: One Foundation Model for Image-Language and Video-Language Tasks | This paper presents OmniVL, a new foundation model to support both image-language and video-language tasks using one universal architecture. It adopts a unified transformer-based visual encoder for both image and video inputs, and thus can perform joint image-language and video-language pretraining. We demonstrate, for... | Accept | After the authors’ rebuttal and long discussion between reviewers and authors, the paper unanimously receives positive rates thanks to reasonable proposed ideas and thorough experiment evaluation. The camera-ready version may need to be updated to fully reflect reviewers’ comments and authors’ answers to them. | train | [
"nYqwlUBKlVI",
"ZIfNK8iuV6",
"w1VRCf9ZP0V",
"TFylbKhtkjC",
"7MYTJ8PiWj",
"0PW8dXYL9JK",
"9W2eiUcEde8",
"rxn7sYxSw2x",
"fsd9trNF8Yb",
"Pf3Wt915eCGW",
"BB1DOws3kxf",
"yaYZtrdpnZ1",
"0PZydnhoyTi",
"h1aMacnU5R74",
"oBjCUagS_Lve",
"gH7PhVI-4Zu",
"vdzHGhjgWgS",
"C3SXGGQbbGf",
"FEENJNLy... | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",... | [
" Dear Reviewer cPTx, thanks for your effort again! We are happy that our rebuttal well addressed your concerns!",
" Thanks for your rebuttal, I appreciated the newly added experiments. \nThe comparison becomes stronger now. My concerns about NoCaps are addressed.\nI have updated my ratings. ",
" Hi cPTx, sorry... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
5,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
5,
4
] | [
"ZIfNK8iuV6",
"rxn7sYxSw2x",
"rxn7sYxSw2x",
"fsd9trNF8Yb",
"rxn7sYxSw2x",
"rxn7sYxSw2x",
"rxn7sYxSw2x",
"yaYZtrdpnZ1",
"Pf3Wt915eCGW",
"hfx1QwdfuGq",
"0PZydnhoyTi",
"gH7PhVI-4Zu",
"vdzHGhjgWgS",
"nLUKAMW89Qp",
"VvtFH5Je5F8",
"FEENJNLySr",
"nLUKAMW89Qp",
"XeXtqYvmEth",
"XeXtqYvmEt... |
nips_2022_MK_130d4Y0 | EcoFormer: Energy-Saving Attention with Linear Complexity | Transformer is a transformative framework for deep learning which models sequential data and has achieved remarkable performance on a wide range of tasks, but with high computational and energy cost. To improve its efficiency, a popular choice is to compress the models via binarization which constrains the floating-poi... | Accept | The paper after rebuttal addresses several of the limitations (mainly lacking positioning in the rich existing litterature) of the first submission. The strength of the paper resides in a holistic approach to the ("yet another") efficient attention mechanism, evaluating and discussing trade-offs between accuracy, compu... | train | [
"b9e4vnBdlPR",
"1qgwimKOVk",
"52Z4Skec2oS",
"N1ceu5aE7vj",
"-rsJyQ48coQ",
"rOlJjXV_Vo",
"_kVcygdjc35",
"K1yx3iLGW0c",
"vQo7xYIcJ3U",
"upxPvn_UeKO",
"NVDIVFp2yBV",
"9KMjYLmH5Ux",
"FDyN9Q7F9wJ",
"mCY_ZVqAaCq",
"7ODSGDPNTTW"
] | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for your feedback and suggestions! We are happy to address your questions and appreciate the valuable comments.",
" Thanks for your feedback and suggestions! We feel glad to address your questions and appreciate the constructive reviews for improving our work.",
" Thank you for the clarifications. I am... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
6,
4
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
5,
4
] | [
"52Z4Skec2oS",
"N1ceu5aE7vj",
"upxPvn_UeKO",
"9KMjYLmH5Ux",
"7ODSGDPNTTW",
"FDyN9Q7F9wJ",
"nips_2022_MK_130d4Y0",
"7ODSGDPNTTW",
"mCY_ZVqAaCq",
"mCY_ZVqAaCq",
"FDyN9Q7F9wJ",
"FDyN9Q7F9wJ",
"nips_2022_MK_130d4Y0",
"nips_2022_MK_130d4Y0",
"nips_2022_MK_130d4Y0"
] |
nips_2022_9uRS5ysgb9 | Zero-Shot Video Question Answering via Frozen Bidirectional Language Models | Video question answering (VideoQA) is a complex task that requires diverse multi-modal data for training. Manual annotation of question and answers for videos, however, is tedious and prohibits scalability. To tackle this problem, recent methods consider zero-shot settings with no manual annotation of visual question-a... | Accept | The submission introduces a zero-shot VQA model that combines frozen video and bidirectional language models by training additional projection and adaptor layers. The method significantly outperforms related previous work that uses only uni-directional language models. While the approach is somewhat incremental technic... | train | [
"CMTc8d1OZR",
"yHcrV_mWLxY",
"ooAxjkEcTqq",
"RdET692ma4X",
"g1B5n0Sd_c",
"nrBbq_Agt_1",
"vCruQ2zvQq",
"9jCdQXk4vcA",
"hkdciq9JU1Q",
"BeuJpAWzhE7",
"eSctdZRa-v"
] | [
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for the detailed response and the extra experiments on the template design and image VQA. After reading all the reviews and discussions, I am happy to support this paper for acceptance.",
" Thanks for the clarification and insights. The additional results and discussions on \"Multiple mask tokens deco... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4
] | [
"vCruQ2zvQq",
"nrBbq_Agt_1",
"9jCdQXk4vcA",
"nips_2022_9uRS5ysgb9",
"eSctdZRa-v",
"eSctdZRa-v",
"BeuJpAWzhE7",
"hkdciq9JU1Q",
"nips_2022_9uRS5ysgb9",
"nips_2022_9uRS5ysgb9",
"nips_2022_9uRS5ysgb9"
] |
nips_2022_qVtbqSwOxy6 | Align then Fusion: Generalized Large-scale Multi-view Clustering with Anchor Matching Correspondences | Multi-view anchor graph clustering selects representative anchors to avoid full pair-wise similarities and therefore reduce the complexity of graph methods. Although widely applied in large-scale applications, existing approaches do not pay sufficient attention to establishing correct correspondences between the anchor... | Accept | All reviewer agree that this paper is innovative and well-written, so I recommend to accept. | train | [
"pjfcPGSA_n",
"KgNmZ-8kr1r",
"om9Mo9QBjc7",
"TZEHN71GfF6",
"RVMFvzkE27N",
"TvPwDkQRS7u",
"ZTZGPnzZxP",
"uVQV8IILv4Z",
"bABADCSpOLG",
"40AV9J-7qRR"
] | [
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for your response. I agree that the idea of recognizing the multi-view anchor alignment problem has been overlooked in the existing literature which can benefit further large-scale multi-view clustering research.\n\nI think the authors have addressed all of my concerns. I will keep my score and suggest ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
5
] | [
"TvPwDkQRS7u",
"nips_2022_qVtbqSwOxy6",
"TZEHN71GfF6",
"RVMFvzkE27N",
"40AV9J-7qRR",
"bABADCSpOLG",
"uVQV8IILv4Z",
"nips_2022_qVtbqSwOxy6",
"nips_2022_qVtbqSwOxy6",
"nips_2022_qVtbqSwOxy6"
] |
nips_2022_dozWFpOJcOD | RLIP: Relational Language-Image Pre-training for Human-Object Interaction Detection | The task of Human-Object Interaction (HOI) detection targets fine-grained visual parsing of humans interacting with their environment, enabling a broad range of applications. Prior work has demonstrated the benefits of effective architecture design and integration of relevant cues for more accurate HOI detection. Howev... | Accept | This paper proposes a free-form relational language-image pretraining for HOI detection which demonstrates advantageous performances on zero-shot and few-shot settings. All reviewers give consistent positive scores after the discussion phase. The authors have added more experiments on COCO, different backbones, and pre... | train | [
"vDnLFsSRHM3",
"o9WY5a6Vkz",
"WPF9Cu-rZbB",
"ZBA-rt9urlX",
"Hqnf52jd34f",
"xmBATmLpk1p",
"-Hu-PVfkLYB6",
"ZjqTlwGsx5X",
"65TjhInlmmE",
"r9EszBO1IME3",
"KHRMn28J_XT",
"iVzx1vmNwZ5",
"FAbYq0yfExo",
"-sVn1hTOTJv",
"mOtc1qJ9ODB",
"RmJBg29XtbWH",
"2GV0YnWQDMi",
"7JUh8jxc6P",
"SiYJ9a79... | [
"author",
"official_reviewer",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
... | [
" Dear Chairs and Reviewers,\n\nAs the discussion period comes to an end, we want to present a brief summary of our rebuttal and discussions with the reviewers for further reference.\n\nFirst of all, we thank all reviewers for their efforts and valuable comments. We are encouraged that the reviewers found RLIP show... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
5,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
5,
4
] | [
"nips_2022_dozWFpOJcOD",
"LAzp5br9CW8",
"xmBATmLpk1p",
"hL0EmVH3E0",
"65TjhInlmmE",
"ZjqTlwGsx5X",
"iVzx1vmNwZ5",
"iVzx1vmNwZ5",
"iVzx1vmNwZ5",
"LAzp5br9CW8",
"RmJBg29XtbWH",
"SiYJ9a797MCO",
"SiYJ9a797MCO",
"7JUh8jxc6P",
"2GV0YnWQDMi",
"459_oI38wZk",
"r0X9ieSkI3G",
"r0X9ieSkI3G",
... |
nips_2022_1tIUqrUuJxx | Dynamic Graph Neural Networks Under Spatio-Temporal Distribution Shift | Dynamic graph neural networks (DyGNNs) have demonstrated powerful predictive abilities by exploiting graph structural and temporal dynamics. However, the existing DyGNNs fail to handle distribution shifts, which naturally exist in dynamic graphs, mainly because the patterns exploited by DyGNNs may be variant with respe... | Accept | The paper addresses spatio-temporal distribution shifts in dynamic graphs by discovering and utilizing invariant patterns, i.e., structures and features whose predictive abilities are stable across distribution shifts. The paper is an early try to address distribution shifts in dynamic graphs, which is an interesting a... | train | [
"OTXJw4S6ZGZ",
"1HLyYoLO9l9",
"KgEybT3zjO",
"HX3HQ0SV8D",
"Be13RQ55AvO",
"7HprgfhVUhu",
"XjgVTbz8wTY",
"PA0Y4Utyceh"
] | [
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" **Q1: The invariant pattern can be further divided into the time dependent and time independent ones.**\n\nA1: Thank you for your comment. We agree that further analyzing and dividing the invariant patterns as you suggest is an interesting idea. As our paper is the first work on studying spatio-temporal distribut... | [
-1,
-1,
-1,
-1,
6,
7,
5,
7
] | [
-1,
-1,
-1,
-1,
4,
4,
2,
4
] | [
"PA0Y4Utyceh",
"XjgVTbz8wTY",
"7HprgfhVUhu",
"Be13RQ55AvO",
"nips_2022_1tIUqrUuJxx",
"nips_2022_1tIUqrUuJxx",
"nips_2022_1tIUqrUuJxx",
"nips_2022_1tIUqrUuJxx"
] |
nips_2022_V0GwAmDclY | Mix and Reason: Reasoning over Semantic Topology with Data Mixing for Domain Generalization | Domain generalization (DG) enables generalizing a learning machine from multiple seen source domains to an unseen target one. The general objective of DG methods is to learn semantic representations that are independent of domain labels, which is theoretically sound but empirically challenged due to the complex mixture... | Accept | Some of the reviewers had concerns about novelty, and one of the reviewers was worried about the care taken in training a baseline. however, another reviewer has a strong positive opinion of the work; and I believe the authors have done a good job in rebuttal making an effort to address the concerns about baselines. ... | train | [
"855brQq01nk",
"hXnb747Gj5",
"VWPk2P2okb",
"dWMpRJCMSss",
"S2Z7cRsjYD",
"h9ubU-rWJTe",
"0bEcBQXYkem",
"xpXcCGcxGjgo",
"BLlgIBSSKvv",
"--9i3pChYHZ",
"mu0MlNypcjl",
"eqp6GICYoGg",
"J-uJR5axuE-E",
"mYc9QRSbGv4",
"AHOInLLhnC4",
"xoOy7AgJQsk",
"_HI031sf3GH",
"6A7Tnlfyq-K",
"Bd4y922KDc... | [
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_... | [
" Dear Reviewer GJ2F,\n\nWe are wondering if you could give some comments and final thoughts about our previous discussion. Please let us know if you have further follow-up discussions. We would be immensely grateful if you could raise the rating to reflect the contributions of our paper. \n\nThank you very much.\n... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
5,
4
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
5
] | [
"dWMpRJCMSss",
"S2Z7cRsjYD",
"dWMpRJCMSss",
"xoOy7AgJQsk",
"h9ubU-rWJTe",
"eqp6GICYoGg",
"mu0MlNypcjl",
"--9i3pChYHZ",
"J-uJR5axuE-E",
"xI84Kufi91h",
"Bd4y922KDc",
"6A7Tnlfyq-K",
"_HI031sf3GH",
"IoGjQ2WD_Eg",
"Hpru00A7Uh6",
"Hpru00A7Uh6",
"3gj03t8gHcw",
"3gj03t8gHcw",
"3gj03t8gHc... |
nips_2022_wOUH1VQ9Rcj | Independence Testing-Based Approach to Causal Discovery under Measurement Error and Linear Non-Gaussian Models | Causal discovery aims to recover causal structures generating the observational data. Despite its success in certain problems, in many real-world scenarios the observed variables are not the target variables of interest, but the imperfect measures of the target variables. Causal discovery under measurement error aims t... | Accept | The paper considers structure recovery when there is a causal DAG on variables X where causal mechanisms are linear but exogenous noise variables are non-Gaussian (similar setting to the one in the standard prior work LiNGAM). However, each variable is not directly observed but through measurement independent noise. Au... | train | [
"wbR9vm2xf9t",
"ghe14-Hi10_",
"QQR9gdMsS6g",
"zgCT06fduYl",
"BYVH4gFJDI1",
"nRsnyraVvXw",
"KYTwdET-oL4",
"9DWllf1-7Oe",
"iM-Jn9WiqYK",
"KTiX5jVnXTd",
"LT6sP6DIKag",
"p-R24Ww0rMt",
"tY2DqqBEXr",
"uGazqMkq2Tr",
"Buk3zAOY4Iy",
"yfcp09bj2sA",
"Bqd-ckTb1Nu",
"uqNkHfH43R4"
] | [
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear Reviewer Pi4t,\n\nOnce again, thanks a lot for reviewing our submission and for the further iterations! We hope we have properly addressed your major concerns. If that is the case, could you please consider increasing your score? (As you kindly mentioned in your comments, you would be happy to increase you... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
2,
3,
4
] | [
"uqNkHfH43R4",
"nips_2022_wOUH1VQ9Rcj",
"Buk3zAOY4Iy",
"nRsnyraVvXw",
"KYTwdET-oL4",
"9DWllf1-7Oe",
"p-R24Ww0rMt",
"KTiX5jVnXTd",
"Buk3zAOY4Iy",
"tY2DqqBEXr",
"tY2DqqBEXr",
"tY2DqqBEXr",
"uqNkHfH43R4",
"Bqd-ckTb1Nu",
"yfcp09bj2sA",
"nips_2022_wOUH1VQ9Rcj",
"nips_2022_wOUH1VQ9Rcj",
... |
nips_2022_AUz5Oig77OS | Efficient Spatially Sparse Inference for Conditional GANs and Diffusion Models | During image editing, existing deep generative models tend to re-synthesize the entire output from scratch, including the unedited regions. This leads to a significant waste of computation, especially for minor editing operations. In this work, we present Spatially Sparse Inference (SSI), a general-purpose technique th... | Accept | This paper was a close call. One reviewer was of the opinion that the paper lacked significant innovations other than fairly obvious sparse processing tricks to make local edits faster. This reviewer did not change their opinion (Borderline Reject) post-rebuttal. Of the other two reviewers, one was at Borderline Accept... | train | [
"rWTWydGgaw",
"e4QLEjG9WQ",
"b7B-8CQrkzi",
"fOdYYZh-eI",
"vRSdHWuv0tG",
"QRWdS8ZiWk",
"lGy0nX-u4gp",
"0QxGREJx-jf",
"GvMCGqQAEQW",
"mbcEGReMDZ7",
"cNv5wyJI6Be",
"4SWB61qGiU_",
"fefQ6Z09uJE",
"3k3NZqlNbo"
] | [
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for your additional comments. In our general response, we highlight our novelty and differences between our work and SBNet. Please do not hesitate to contact us if you have additional questions. Thank you for your time again!",
" Thanks for your additional comments. In our general response, we highlight ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
6,
4
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
4
] | [
"vRSdHWuv0tG",
"fOdYYZh-eI",
"lGy0nX-u4gp",
"QRWdS8ZiWk",
"mbcEGReMDZ7",
"0QxGREJx-jf",
"cNv5wyJI6Be",
"3k3NZqlNbo",
"fefQ6Z09uJE",
"4SWB61qGiU_",
"nips_2022_AUz5Oig77OS",
"nips_2022_AUz5Oig77OS",
"nips_2022_AUz5Oig77OS",
"nips_2022_AUz5Oig77OS"
] |
nips_2022_5kThooa07pf | Subsidiary Prototype Alignment for Universal Domain Adaptation | Universal Domain Adaptation (UniDA) deals with the problem of knowledge transfer between two datasets with domain-shift as well as category-shift. The goal is to categorize unlabeled target samples, either into one of the "known" categories or into a single "unknown" category. A major problem in UniDA is negative trans... | Accept | This submission deals with universal domain adaptation for object recognition. The authors propose to extend existing strategies with an original and effective complementary strategy, thus achieving SOTA performance in this context. Their first proposal aims to align domains while avoiding the risk of negative-transfer... | train | [
"oFUkIf2C8Mz",
"xUJ9-ffo2Lf",
"AzH0LYJA4pp",
"EOnkqI0Q2Xx",
"RwR6AwK7LvI",
"XiJKydjMyyT",
"KTuY23ziFE",
"Rka9DO0a28F4",
"f1vBAiKrTl",
"fJ5hHKyQi4",
"TH2NZ5AtnNq",
"xiJvVJFws9o",
"8jxNnJDPad"
] | [
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for the response. This clarifies my question on NTR and Fig 2. It would be good to improve the current presentation of NTR/Fig.2 by incorporating this clarification discussion. \n\nI appreciate the efforts of adding extra experiments on additional ablation and backbone. However, in my opinion, the rebuttal... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
4,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
4
] | [
"Rka9DO0a28F4",
"f1vBAiKrTl",
"EOnkqI0Q2Xx",
"RwR6AwK7LvI",
"KTuY23ziFE",
"nips_2022_5kThooa07pf",
"8jxNnJDPad",
"xiJvVJFws9o",
"TH2NZ5AtnNq",
"nips_2022_5kThooa07pf",
"nips_2022_5kThooa07pf",
"nips_2022_5kThooa07pf",
"nips_2022_5kThooa07pf"
] |
nips_2022_C2o5DeL_8L1 | Generative Status Estimation and Information Decoupling for Image Rain Removal | Image rain removal requires the accurate separation between the pixels of the rain streaks and object textures. But the confusing appearances of rains and objects lead to the misunderstanding of pixels, thus remaining the rain streaks or missing the object details in the result. In this paper, we propose SEIDNet equipp... | Accept | The paper has received positive reviews. There was substantial discussion, and the authors are strongly encouraged to include the clarifications they made rot the final copy, as well as a more extensive discussion of the limitations and possible danger of overfitting.
| train | [
"4QRMLZoF8Nf",
"BgvC5ClVoo",
"bM9uyohn3_O",
"2a1gdcyadCg",
"zerIWTSmoQ2",
"VQ7-6K-JGd",
"UDgBwNfA_x",
"L2iiAa4qSpE",
"GTIh28WpWYC",
"Ag-w-XT0tbr",
"H1r7ZgrZrJv",
"8nE2lBL3dMrW",
"RvlUfkOPTHo",
"AloCWlACimr",
"qbGMjmMj4gSy",
"jI_9aZuTOj",
"gJyq0rWr8Gb",
"gsBSsL0cktU",
"Jz3qr1PsWiA... | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official... | [
" Thanks for your new comment again!\n\n**Is there any intuitive reason for the proposed linear blending formulation of K in EQ7? How does the K_o component (with/without K_o) impact the performance? Have you ever tried other formulations for K?**\n\nThough Eq. 7 is a linear blending form of kernels, the learning o... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
5,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
4,
3,
3
] | [
"bM9uyohn3_O",
"2a1gdcyadCg",
"jI_9aZuTOj",
"qbGMjmMj4gSy",
"VQ7-6K-JGd",
"UDgBwNfA_x",
"H1r7ZgrZrJv",
"yHV12xyKbBf",
"nFOWanRn1Tf",
"u33v0wrZ5s6",
"Jz3qr1PsWiA",
"yHV12xyKbBf",
"WrZIzoaM5jI",
"nFOWanRn1Tf",
"u33v0wrZ5s6",
"u33v0wrZ5s6",
"nips_2022_C2o5DeL_8L1",
"yHV12xyKbBf",
"W... |
nips_2022_xl39QEYiB-j | Embodied Scene-aware Human Pose Estimation | We propose embodied scene-aware human pose estimation where we estimate 3D poses based on a simulated agent's proprioception and scene awareness, along with external third-person observations. Unlike prior methods that often resort to multistage optimization, non-causal inference, and complex contact modeling to estim... | Accept | The submission initially received mixed reviews. After rebuttal, all reviewers felt their concerns reasonably addressed and recommended acceptance (though one didn't update the score). The AC agrees. The authors are encouraged to revise the paper accordingly. | train | [
"wlQx5Hv8vO",
"EWxdfi3HkRb",
"3H8Ew3i5aW",
"fTJ1P47rN9",
"7SsnyynOubE",
"HiaR2QimTh0",
"5qpmtk4eKqN",
"857o7k4RBv",
"56siKCpoiLm",
"v9GjVoG0eO",
"gxDgjW42ROD",
"0yNMNk22zsq0",
"CuJQUSKM0rv",
"YJBxFeWZnUS",
"gCyMjZFnoI7",
"cVRMx7QTUy",
"b9WKbfEp1CF",
"fTq_pmTWJLA"
] | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you so much for the updated review and response! We are super glad that our reply has answered some of your concerns and improved your impression of our work. Feel free to ask us anything if there are any additional questions!\n\nAlso, if we may ask, could you kindly update the ratings in the official revie... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
5,
4,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
4,
4
] | [
"EWxdfi3HkRb",
"0yNMNk22zsq0",
"gCyMjZFnoI7",
"cVRMx7QTUy",
"b9WKbfEp1CF",
"fTq_pmTWJLA",
"nips_2022_xl39QEYiB-j",
"gCyMjZFnoI7",
"cVRMx7QTUy",
"gxDgjW42ROD",
"b9WKbfEp1CF",
"CuJQUSKM0rv",
"YJBxFeWZnUS",
"fTq_pmTWJLA",
"nips_2022_xl39QEYiB-j",
"nips_2022_xl39QEYiB-j",
"nips_2022_xl39... |
nips_2022_NaW6T93F34m | "Lossless" Compression of Deep Neural Networks: A High-dimensional Neural Tangent Kernel Approach | Modern deep neural networks (DNNs) are extremely powerful; however, this comes at the price of increased depth and having more parameters per layer, making their training and inference more computationally challenging.
In an attempt to address this key limitation, efforts have been devoted to the compression (e.g., sp... | Accept | In the paper, the authors provide theorems that establish that for GMM input data, the NTK matrices of dense and quantized DNNs have the same eigenspectra in the asymptotic limit of high input data dimension and sample size. These results motivate network compression algorithms which demonstrate good empirical perform... | train | [
"IXLVzmIH6g8",
"yy91Fv9CVxP",
"0N4xOCe-_7",
"ZCTp-IAWGSJ",
"2kdD1kHCEu",
"F62y0GUFAy",
"STPC6544o5W",
"zkrDwDS-cBd"
] | [
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" I'd like to thank the authors for their response. It addressed all my questions. Especially, the performance-complexity tradeoff in Figure 3 is very helpful. I decided to keep my rating (7: accept).",
" We thank the reviewer for his/her positive support and for the valuable insights shared in the constructive c... | [
-1,
-1,
-1,
-1,
-1,
6,
5,
7
] | [
-1,
-1,
-1,
-1,
-1,
3,
4,
3
] | [
"2kdD1kHCEu",
"STPC6544o5W",
"F62y0GUFAy",
"F62y0GUFAy",
"zkrDwDS-cBd",
"nips_2022_NaW6T93F34m",
"nips_2022_NaW6T93F34m",
"nips_2022_NaW6T93F34m"
] |
nips_2022_6RoAxmwj0L2 | DaDA: Distortion-aware Domain Adaptation for Unsupervised Semantic Segmentation | Distributional shifts in photometry and texture have been extensively studied for unsupervised domain adaptation, but their counterparts in optical distortion have been largely neglected. In this work, we tackle the task of unsupervised domain adaptation for semantic image segmentation where unknown optical distortion ... | Accept | This paper proposes a new segmentation method with geometric insight to deal with the distortions. As pointed out by our reviewers, this paper is featured with important practical value, clear problem definition, and interesting mathematical insight. During the rebuttal phase, most of the reviewers confirmed their sup... | test | [
"2usz1Iy0GZ",
"kLfAjZv08pD",
"g6-jD-Y8SY3",
"iVsVTBe-u-8",
"ZkXp99NEVa",
"YSOJdSaN4t2",
"mPmdStNPsz-",
"IFIbszCvCJN",
"eN-yfFh25M0",
"uUQG0prGnPi",
"ew14s2TDnI",
"KHblGrI1xoq",
"RmBGanDRVtz",
"AKd3LobaIMp",
"RuxFDBQMHBh"
] | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We appreciate your favorable reviews. According to the reviewer's suggestions, we will add a detailed explanation in the Discussion and Conclusion (currently Conclusion section in the manuscript) section discussing the possibility of extending our methods, to the general domain adaptation methodology.\nWe will de... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
3,
4
] | [
"iVsVTBe-u-8",
"g6-jD-Y8SY3",
"KHblGrI1xoq",
"uUQG0prGnPi",
"YSOJdSaN4t2",
"mPmdStNPsz-",
"RuxFDBQMHBh",
"AKd3LobaIMp",
"AKd3LobaIMp",
"AKd3LobaIMp",
"RmBGanDRVtz",
"RmBGanDRVtz",
"nips_2022_6RoAxmwj0L2",
"nips_2022_6RoAxmwj0L2",
"nips_2022_6RoAxmwj0L2"
] |
nips_2022_AlgbeSuE1lx | Coded Residual Transform for Generalizable Deep Metric Learning | A fundamental challenge in deep metric learning is the generalization capability of the feature embedding network model since the embedding network learned on training classes need to be evaluated on new test classes. To address this challenge, in this paper, we introduce a new method called coded residual transform (... | Accept | This paper proposes a coded residual transform for deep metric learning, aiming to improve the generalization ability of metric learning to unseen classes. Four expert reviewers assessed this paper, with preliminary reviews at odds. After author rebuttal, some reviewers acknowledged the rebuttal by increasing the score... | train | [
"nnWBwDzavyp",
"5iq1rMqBaPp",
"X5MBPsKA4yd",
"sdZCGbSh8Tf",
"D5TrfcIVTY9",
"SC70x8mq18e",
"7M_LY0cjZZ",
"id7VFp3f6Y0",
"s0ATuTBWe3T",
"t6L395q_4WT",
"K747PMs2fm",
"mzAUlJFng8U",
"pAa6aGXQvFz",
"yFX5wB0tId_",
"NKgm4-zf9uc"
] | [
"official_reviewer",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for the thoughtful response to my review and others. I will keep my current rating and hope to see this paper accepted. Nice work authors!",
" Dear Reviewer, \n\nWe really appreciate your kind reply!\n\nSorry for the confusion about the motivation part. In the original paper, we used the **toy example... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
6,
8,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
3,
5
] | [
"s0ATuTBWe3T",
"sdZCGbSh8Tf",
"D5TrfcIVTY9",
"K747PMs2fm",
"t6L395q_4WT",
"7M_LY0cjZZ",
"id7VFp3f6Y0",
"NKgm4-zf9uc",
"yFX5wB0tId_",
"pAa6aGXQvFz",
"mzAUlJFng8U",
"nips_2022_AlgbeSuE1lx",
"nips_2022_AlgbeSuE1lx",
"nips_2022_AlgbeSuE1lx",
"nips_2022_AlgbeSuE1lx"
] |
nips_2022_ZVe_WeMold | S-Prompts Learning with Pre-trained Transformers: An Occam’s Razor for Domain Incremental Learning | State-of-the-art deep neural networks are still struggling to address the catastrophic forgetting problem in continual learning. In this paper, we propose one simple paradigm (named as S-Prompting) and two concrete approaches to highly reduce the forgetting degree in one of the most typical continual learning scenarios... | Accept | This paper adopts the prompt learning idea into continual learning to tackle the problem of domain incremental learning. The proposed approach is clear, the writing is easy to follow. The experiment is convincing. | train | [
"aksVSJo4sNX",
"xXnmBu9kF6y",
"fQGVbluGqzp",
"q7l6hc8e6bL",
"I7dxnysR8lN",
"tJ-1SoaJAuY",
"jPrTj8aczV",
"mG2WgYO8JIYP",
"W4bwHiTNmX",
"jfvoJJ2rlp4-",
"XCiMWx72JdE",
"IUXGTrCY-82",
"5a8MvLkaIJw",
"mT0j2wB1cJ9",
"sPqkMChnQS8",
"1vElwhZa_de"
] | [
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" \nThanks for making the further comment. The reason why we did not provide the performance of the base-ViT for DyTox on CDDB in the main paper is three-fold. Firstly, the original DyTox paper [14] suggests using a more advanced ViT (i.e., base-ConViT [15]), and thus we directly evaluate its officially implemented... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
6,
4,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
3,
5
] | [
"q7l6hc8e6bL",
"fQGVbluGqzp",
"tJ-1SoaJAuY",
"I7dxnysR8lN",
"1vElwhZa_de",
"1vElwhZa_de",
"sPqkMChnQS8",
"sPqkMChnQS8",
"mT0j2wB1cJ9",
"5a8MvLkaIJw",
"5a8MvLkaIJw",
"nips_2022_ZVe_WeMold",
"nips_2022_ZVe_WeMold",
"nips_2022_ZVe_WeMold",
"nips_2022_ZVe_WeMold",
"nips_2022_ZVe_WeMold"
] |
nips_2022_Nay_rOB-dZv | Fairness Reprogramming | Despite a surge of recent advances in promoting machine Learning (ML) fairness, the existing mainstream approaches mostly require training or finetuning the entire weights of the neural network to meet the fairness criteria. However, this is often infeasible in practice for those large-scale trained models due to larg... | Accept | Overall the reviews are more or less positive towards weak accept. While there are some remaining concerns (e.g., a wording in the abstract), I think many of the raised concerns are addressed properly and some of them are checked. Hence, I believe this paper is worth being published. | train | [
"LtynAqCKHMy",
"5x8f7gUiSmQ",
"Q1HQ6XAbyCP",
"BTXJ_OzD4-s",
"jMOORhBbK6V",
"BayFPPow41_",
"oRN30sbMUz4",
"vmSYix3tA_j",
"DCIvhAxBZhX",
"0z68iCryE3W",
"Poxf-QqS3r3",
"h9g2DSFxE4",
"DTF8LyaMiyu",
"KBgjXE1o_dJ",
"uGAFzuzabti",
"1cFuszm2Ot2",
"U1hwY0v36Ep",
"1yrLfnhXXf8"
] | [
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" I see what you mean. In my mind, the obtaining/applying distinction isn't especially meaningful in terms of what level of access is assumed for the proposed method. For example you said \"consider a black-box NLP model\" not \"consider a black-box NLP model where the fairness trigger has previously been obtained\... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
5,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
2
] | [
"5x8f7gUiSmQ",
"Q1HQ6XAbyCP",
"0z68iCryE3W",
"Poxf-QqS3r3",
"U1hwY0v36Ep",
"1yrLfnhXXf8",
"Poxf-QqS3r3",
"1yrLfnhXXf8",
"U1hwY0v36Ep",
"Poxf-QqS3r3",
"uGAFzuzabti",
"1yrLfnhXXf8",
"U1hwY0v36Ep",
"U1hwY0v36Ep",
"1cFuszm2Ot2",
"nips_2022_Nay_rOB-dZv",
"nips_2022_Nay_rOB-dZv",
"nips_2... |
nips_2022_Wk-4Tp-gPpv | DeepTOP: Deep Threshold-Optimal Policy for MDPs and RMABs | We consider the problem of learning the optimal threshold policy for control problems. Threshold policies make control decisions by evaluating whether an element of the system state exceeds a certain threshold, whose value is determined by other elements of the system state. By leveraging the monotone property of thres... | Accept | The paper considers a subset of dynamic problems, in which the optimal policy is a threshold-policy. The authors use this attribute to formulate tailored off-policy actor-critic algorithms, for both MDPs and RMABs which are gradient-based, so can utilize neural networks. They empirically compare their method to SOTA me... | train | [
"qAe_fhsq1X",
"5oI5LiEo1Hu",
"kamPV9LE01p",
"WqbAQDp5lv",
"eAbmRpC9JAP",
"-zsIjXX3B7o",
"MhxaGhZAmsr",
"dasj9-QtSe2",
"eLB9DSYDy1",
"yWi5p3H1JJd",
"rN4GoSrG_N9g",
"zuKQjTy8M3s",
"A1pXy6wyFil",
"SteInCClNqx",
"xHfCq08hMjb"
] | [
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for the comments. For the first question:\n\nWe agree that the following statements are true:\n1. If all three actor-critic algorithms use the same rule for updating the actor network, then they become the same and would have exactly the same performance.\n2. If we run the algorithms for a much longer t... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
5,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
4,
3
] | [
"5oI5LiEo1Hu",
"-zsIjXX3B7o",
"dasj9-QtSe2",
"eLB9DSYDy1",
"xHfCq08hMjb",
"MhxaGhZAmsr",
"SteInCClNqx",
"A1pXy6wyFil",
"yWi5p3H1JJd",
"rN4GoSrG_N9g",
"zuKQjTy8M3s",
"nips_2022_Wk-4Tp-gPpv",
"nips_2022_Wk-4Tp-gPpv",
"nips_2022_Wk-4Tp-gPpv",
"nips_2022_Wk-4Tp-gPpv"
] |
nips_2022_DRckHIGk8qw | GAMA: Generative Adversarial Multi-Object Scene Attacks | The majority of methods for crafting adversarial attacks have focused on scenes with a single dominant object (e.g., images from ImageNet). On the other hand, natural scenes include multiple dominant objects that are semantically related. Thus, it is crucial to explore designing attack strategies that look beyond learn... | Accept | The authors proposed the first multi-object generative attack, GAMA, which utilizes the vision-language model CLIP as an attacker's tool in the training of the generator to enhance the transferability across different data distributions.
All four reviewers recognize that this paper is well-written and easy to follow. ... | train | [
"QyfrIm3BQsQ",
"CxD35dDmZJ",
"3mXohj1tR32",
"f57evnvBH0K",
"Gc3-B6yxEke",
"1MSqlPHSzsD",
"yYOOct9j0HC",
"M7fbwTR7F7K",
"2UlfMjkTtu",
"Sqn9nNWBfIZ",
"OfEEY6S44KP",
"JFtHaGgFVKO",
"SykELBQK34F",
"mUBB6Yiyouj",
"qYPh2wQ3Lw1"
] | [
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for taking the time to read our response. We are happy to see your concerns are addressed.\n\nBest Wishes,\n\nAuthors",
" Thanks for the response from the authors. It is great to see the method performs well on single-object datasets. My concerns are addressed.",
" Thank you for taking the time to r... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
6,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
5,
3
] | [
"CxD35dDmZJ",
"2UlfMjkTtu",
"f57evnvBH0K",
"M7fbwTR7F7K",
"nips_2022_DRckHIGk8qw",
"Sqn9nNWBfIZ",
"qYPh2wQ3Lw1",
"mUBB6Yiyouj",
"SykELBQK34F",
"JFtHaGgFVKO",
"nips_2022_DRckHIGk8qw",
"nips_2022_DRckHIGk8qw",
"nips_2022_DRckHIGk8qw",
"nips_2022_DRckHIGk8qw",
"nips_2022_DRckHIGk8qw"
] |
nips_2022_NjP18IbKKlX | RecursiveMix: Mixed Learning with History | Mix-based augmentation has been proven fundamental to the generalization of deep vision models. However, current augmentations only mix samples from the current data batch during training, which ignores the possible knowledge accumulated in the learning history. In this paper, we propose a recursive mixed-sample learni... | Accept | The manuscript has been reviewed by five reviewers with ratings of 4,5,6,7,7. The reviewers are in general happy with the contributions, novelty, experimental validation, and mostly recommended acceptance. The AC agrees with the majority vote and would like to recommend acceptance. Congratulations!
| train | [
"TqJI5jxnrkW",
"n8N1RPT6bg",
"iYk5iWcDxYc",
"2Zkcxg9zL5o",
"DvVYBojwvOS",
"iN-PRT1Ju_T",
"jTfOR9eypO",
"VeHVVEm1XJ",
"AhPMzMAaSZ",
"rA-BpWlq17",
"wquMxksVPLm",
"1jWiaCzAw9J",
"jJZ4y760R2E",
"bnP5qpblAct",
"9DE_H4KUnG2",
"uSJSXFHPgh",
"_1p5Mi4FO64"
] | [
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks for the response, which has addressed my concerns. Thus, I won't change my rating.",
" After careful consideration, I decided to increase my rating by one score. \nI hope that our discussion could be carefully resolved in the final revision.\n\nBest",
" Thanks for the clarification. I understand the \"... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
7,
5,
4,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
5,
4
] | [
"1jWiaCzAw9J",
"iYk5iWcDxYc",
"2Zkcxg9zL5o",
"DvVYBojwvOS",
"wquMxksVPLm",
"rA-BpWlq17",
"VeHVVEm1XJ",
"uSJSXFHPgh",
"_1p5Mi4FO64",
"9DE_H4KUnG2",
"bnP5qpblAct",
"jJZ4y760R2E",
"nips_2022_NjP18IbKKlX",
"nips_2022_NjP18IbKKlX",
"nips_2022_NjP18IbKKlX",
"nips_2022_NjP18IbKKlX",
"nips_2... |
nips_2022_gc87Cs_V9qR | Differentiable Analog Quantum Computing for Optimization and Control | We formulate the first differentiable analog quantum computing framework with specific parameterization design at the analog signal (pulse) level to better exploit near-term quantum devices via variational methods. We further propose a scalable approach to estimate the gradients of quantum dynamics using a forward pass... | Accept | The paper proposes a differentiable programming framework for analog quantum computing with a specialized forward scheme based on Monte-Carlo sampling to get estimates of gradients. This idea is an exciting avenue for research to broaden the applicability of quantum computing to practical machine learning and computati... | train | [
"Nosf0UqGoFu",
"2sB6f-VTeQJ",
"KexhInyF2_u",
"41hA4kUcZaO",
"9IP-e8vZrV",
"bs40kD7u1A4",
"PjeXpIM0D6C",
"VNUp8riLOKj",
"cLkf-KqL1UY",
"42uz8HJ-_Oo",
"nGw5cfNhaf3",
"0ibFSo1WjnG",
"4cvqUvIhcEN",
"IqQVqwyw4DW"
] | [
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" I thank the authors for the detailed response, which solve my concerns to some extents. I increase my score accordingly. Param shift rule only applies to gates whose unitary matrix has structured eigenvalues such as +1 and -1. It would be good for the authors to explain more on the limitations of the parameter sh... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
8,
8,
5,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
5,
1
] | [
"2sB6f-VTeQJ",
"VNUp8riLOKj",
"41hA4kUcZaO",
"cLkf-KqL1UY",
"nips_2022_gc87Cs_V9qR",
"IqQVqwyw4DW",
"4cvqUvIhcEN",
"4cvqUvIhcEN",
"0ibFSo1WjnG",
"nGw5cfNhaf3",
"nips_2022_gc87Cs_V9qR",
"nips_2022_gc87Cs_V9qR",
"nips_2022_gc87Cs_V9qR",
"nips_2022_gc87Cs_V9qR"
] |
nips_2022_QXLue5WoSBE | NeuPhysics: Editable Neural Geometry and Physics from Monocular Videos | We present a method for learning 3D geometry and physics parameters of a dynamic scene from only a monocular RGB video input. To decouple the learning of underlying scene geometry from dynamic motion, we represent the scene as a time-invariant signed distance function (SDF) which serves as a reference frame, along with... | Accept | After rebuttal the new version of the paper reads much better and all reviewers were positive, despite some remaining criticisms. Hence the paper should be accepted. | test | [
"ebzMgFLDRw-",
"wQl5Jw4fiE5",
"4BX9OnyYjuO",
"hykuF0AXK0w",
"WIgwaLlI-QY",
"bcIqCjZIGIS",
"24rXIAozY8D",
"lLbKWHX2_xw",
"EHobpGn8b4t",
"jj-rab_nd7l",
"nYnifai3D4L",
"FKthJJInRDZ",
"Y4jGMoyxjSG",
"3toCMXd--Zj",
"RO3fZkQ-MO",
"bGGBmePHRK",
"uESZGMEi6kz"
] | [
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thanks very much for your constructive feedback and for kindly increasing the score! We will switch the order of the figures. More discussions and failure cases about our experience of connecting the rendering and simulation modules will also be included.\n\nWe actually found that sequential training instead of j... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
5,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
3,
4
] | [
"4BX9OnyYjuO",
"nips_2022_QXLue5WoSBE",
"bGGBmePHRK",
"WIgwaLlI-QY",
"lLbKWHX2_xw",
"3toCMXd--Zj",
"uESZGMEi6kz",
"uESZGMEi6kz",
"bGGBmePHRK",
"bGGBmePHRK",
"RO3fZkQ-MO",
"RO3fZkQ-MO",
"RO3fZkQ-MO",
"nips_2022_QXLue5WoSBE",
"nips_2022_QXLue5WoSBE",
"nips_2022_QXLue5WoSBE",
"nips_2022... |
nips_2022_17KCLTbRymw | SGAM: Building a Virtual 3D World through Simultaneous Generation and Mapping | We present simultaneous generation and mapping (SGAM), a novel 3D scene generation algorithm. Our goal is to produce a realistic, globally consistent 3D world on a large scale. Achieving this goal is challenging and goes beyond the capacities of existing 3D generation or video generation approaches, which fail to scale... | Accept | The paper explores generation of a volumetric (voxel map via hashing, a la KinectFusion et seq) from a sequence of 2D images. This is achieved by synthesizing sensor images, and feeding them into a mapping module (like KinectFusion). As the reviewers note, this is an interesting goal, and the approach is reasonable, ... | val | [
"hix-9MqaCIp",
"GSa13V8dXXs",
"R0aJ5pGnN2e",
"VFt-hKF_tLJ",
"84d8L3LSx7u",
"QPdKLA5aWdM",
"y-JwMIVf4YT",
"zAC5VWhLkNr",
"1SvEP7yVVYz"
] | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We thank Reviewer QFY4 for the prompt response! Please see the clarifications below.\n\n**Wordings:** We agree with Reviewer QFY4 that the wording is confusing. We will change from \"*accurate depth*\" to \"*complete geometry*\". Thanks for pointing it out!\n\n**Further clarifications on geometry requirement:** I... | [
-1,
-1,
-1,
-1,
-1,
-1,
5,
8,
3
] | [
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
5
] | [
"GSa13V8dXXs",
"84d8L3LSx7u",
"1SvEP7yVVYz",
"zAC5VWhLkNr",
"y-JwMIVf4YT",
"nips_2022_17KCLTbRymw",
"nips_2022_17KCLTbRymw",
"nips_2022_17KCLTbRymw",
"nips_2022_17KCLTbRymw"
] |
nips_2022_2nWUNTnFijm | Learning Substructure Invariance for Out-of-Distribution Molecular Representations | Molecule representation learning (MRL) has been extensively studied and current methods have shown promising power for various tasks, e.g., molecular property prediction and target identification. However, a common hypothesis of existing methods is that either the model development or experimental evaluation is mostly... | Accept | All reviewers agreed that this paper should be accepted because of the strong author response during the rebuttal phase. Specifically the reviewers appreciated the motivation of the paper, its clarity, and added explanation and experiments included during the rebuttal. Authors: please carefully revise the manuscript ba... | test | [
"R-Z4y7hmRJ5",
"BXRwhl1nzs",
"pZpvRrE6qEb",
"Gqcgq5F8vuZ",
"UkgkyBZZkTt",
"tCIlpBhEJz-",
"XD5NyOQKoxy",
"gHpGy2VpTmf",
"2kGjKTP0kg",
"rpF918npdDM",
"DKAwl6Nbs8",
"iY4zfZyraVT",
"1zPkExBJjSl",
"_4BFN_Q9d_v",
"4VPjzFH-936"
] | [
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Dear Reviewers,\n\nThanks again for your valuable comments and nice suggestions. We are still sincerely looking forward to your feedbacks. It is really a good chance for us to engage in the discussion to help us improve this paper.\n\nSincerely, Authors",
" Dear Reviewer MgvC,\n\nThanks again for your time and ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
5,
7
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
2,
5
] | [
"nips_2022_2nWUNTnFijm",
"XD5NyOQKoxy",
"nips_2022_2nWUNTnFijm",
"XD5NyOQKoxy",
"nips_2022_2nWUNTnFijm",
"4VPjzFH-936",
"gHpGy2VpTmf",
"2kGjKTP0kg",
"rpF918npdDM",
"_4BFN_Q9d_v",
"iY4zfZyraVT",
"1zPkExBJjSl",
"nips_2022_2nWUNTnFijm",
"nips_2022_2nWUNTnFijm",
"nips_2022_2nWUNTnFijm"
] |
nips_2022_PYnSpt3jAz | Lethal Dose Conjecture on Data Poisoning | Data poisoning considers an adversary that distorts the training set of machine learning algorithms for malicious purposes. In this work, we bring to light one conjecture regarding the fundamentals of data poisoning, which we call the Lethal Dose Conjecture. The conjecture states: If $n$ clean training samples are need... | Accept | The reviewers agree that this work proposes an interesting conjecture which is likely to inspire further research.
Congrats!
During the discussion period the following two points were raised by the reviewers:
- The paper should emphasize more strongly that this is just a conjecture (I at minimum have doubts have how ... | train | [
"oDjLZazbYj",
"7tBM42zTi65",
"tevPFoOfuP",
"GYrTIJ6Hzz-",
"R3HNHAd-4Vw",
"iADEwrvxXUU",
"9KutH6bzTQW",
"snmEK91-UZc",
"plp52xoxJVo",
"LZsCF016AwM",
"KG4CSzISFQI",
"h8LBeC6_W7f",
"eQ6JD-5qMW",
"RjUsKWmQRV_",
"M3eLJ5nIbsd",
"EfIevcvvWwp",
"-6u3ZObdlBk",
"AXwHg8sPbPX",
"IpfL2hJEPE7"... | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
... | [
" Thank you very much for this clear, detailed and interesting discussion.\nIn my opinion, the verbal description of the score 6 is most suitable for the paper.\nOverall I think that the paper contains significant scientific contributions to the field, and as such will be a good contribution to NeurIPS.",
" Thank... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
7,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
3,
5,
4
] | [
"GYrTIJ6Hzz-",
"R3HNHAd-4Vw",
"iADEwrvxXUU",
"9KutH6bzTQW",
"snmEK91-UZc",
"M3eLJ5nIbsd",
"LZsCF016AwM",
"h8LBeC6_W7f",
"nips_2022_PYnSpt3jAz",
"KG4CSzISFQI",
"Xwm2zsO3QNj",
"IpfL2hJEPE7",
"RjUsKWmQRV_",
"AXwHg8sPbPX",
"EfIevcvvWwp",
"-6u3ZObdlBk",
"nips_2022_PYnSpt3jAz",
"nips_202... |
nips_2022_VvOcK2DGM7G | Unsupervised Causal Generative Understanding of Images | We present a novel framework for unsupervised object-centric 3D scene understanding that generalizes robustly to out-of-distribution images. To achieve this, we design a causal generative model reflecting the physical process by which an image is produced, when a camera captures a scene containing multiple objects. Thi... | Accept | This paper proposes a NERF-based object-centric VAE generative model, which it argues is a more "causal" generative model than prior attempts. While the approach is somewhat elaborate, it is described well, and involves a novel and quite well-motivated combination of several previously proposed components.
The MCMC sa... | train | [
"IBeOwj3IHvz",
"Z9l6BVvnT_M",
"9su8fnNTGN",
"06TBGz62iLh",
"K2qjTWFxTd",
"eq4XX5t5H1a",
"_wZiZc6J6cl",
"y3Px5v_VW2e",
"inuFxkDPL6e",
"leYYxpqzrrS",
"zxRZn7uIxsL",
"tDW8W6O3RSf",
"2ie0EkiE1zMy",
"QgRhAR4IN0x",
"kl5fJ5Sq2kt",
"26uApj6gTJQ",
"mR-egRdUESP",
"fyK-4_LBOVS",
"ULhH2dVbFd... | [
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"officia... | [
" Thank you once again for your review, we are glad to hear we have resolved your concerns.\n\n> by \"analysis\" I meant more trying to understand how it does that - for example - a simple latent traversal may shed some light on what structures this latent holds.\n\nThanks for the suggestion, we agree this would be... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
7,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
3,
5,
4
] | [
"y3Px5v_VW2e",
"ig5EYClfIUv",
"inuFxkDPL6e",
"K2qjTWFxTd",
"eq4XX5t5H1a",
"nips_2022_VvOcK2DGM7G",
"inuFxkDPL6e",
"2ie0EkiE1zMy",
"leYYxpqzrrS",
"zxRZn7uIxsL",
"tDW8W6O3RSf",
"ULhH2dVbFdo",
"QgRhAR4IN0x",
"VJCcC86eb02",
"26uApj6gTJQ",
"KoUY902ctYs",
"fyK-4_LBOVS",
"ig5EYClfIUv",
... |
nips_2022_PCQyUvAmKs | Don't Pour Cereal into Coffee: Differentiable Temporal Logic for Temporal Action Segmentation | We propose Differentiable Temporal Logic (DTL), a model-agnostic framework that introduces temporal constraints to deep networks. DTL treats the outputs of a network as a truth assignment of a temporal logic formula, and computes a temporal logic loss reflecting the consistency between the output and the constraints. W... | Accept | This paper introduces an approach for incorporating declarative temporal constraints in the training of temporal action segmentation models, in a model-agnostic fashion. Reviewers generally appreciated the proposed approach, but questioned the scalability and generalizability of the constraint curation process and aske... | train | [
"u_PJi06cWh",
"sFBqxrGwVDy",
"34YbgfqKI8Q",
"4o8uq9FFN6l",
"kFCqS9-RH5L",
"wmr7N_o4EC",
"Bp2StD1PttI",
"BSoC_N_0FA",
"MjW28xif1G",
"B3UK9bTAEYw",
"ctdFheZXBVE",
"zT2i59fW39",
"ULmopOvNkej"
] | [
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for the follow-up comment and for recognizing our work. Following the comment, we will update the results under the correct setting in future revisions. ",
" Thank you very much for the comprehensive response.\n\nI think curating constraints from only the training annotations is the correct way here, ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
7,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
5
] | [
"sFBqxrGwVDy",
"BSoC_N_0FA",
"4o8uq9FFN6l",
"wmr7N_o4EC",
"wmr7N_o4EC",
"ULmopOvNkej",
"BSoC_N_0FA",
"zT2i59fW39",
"ctdFheZXBVE",
"nips_2022_PCQyUvAmKs",
"nips_2022_PCQyUvAmKs",
"nips_2022_PCQyUvAmKs",
"nips_2022_PCQyUvAmKs"
] |
nips_2022_tIqzLFf3kk | Rank Diminishing in Deep Neural Networks | The rank of neural networks measures information flowing across layers. It is an instance of a key structural condition that applies across broad domains of machine learning. In particular, the assumption of low-rank feature representations led to algorithmic developments in many architectures. For neural networks, how... | Accept | This paper studied the "rank" of neural networks and showed that deeper network in general will have lower rank. The paper did a detailed empirical study on network rank, as well as some theoretical insights on why rank is likely to decrease as the network becomes deeper, and how the rank decrease can change with or wi... | train | [
"Ktf6ANaq35",
"O0hByKwLsYv",
"NbtfIuV3ttP",
"Vr00movCt16",
"m-3F5lnwJIR",
"p3EkXL3cyMG",
"dHDJMvNsYJj",
"A_u3ofv665Z",
"l3mv-7mVHH2",
"1oh6w5xKLfs",
"WNbTK4UgSzi",
"O18hIUDvph8",
"yMd27bY6XN",
"3s37zZZpxdO",
"6o35yk8l01T",
"9GpuphXn_ZQ",
"FGa8IibP0Y1",
"UJMt8Qm8I0R",
"OKKaDv7xCU6... | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" I thank the authors for their clear and detailed response. This has resolved some of the potential misgivings I had regarding the relationship to other work, and I am happy to increase my score. ",
" I thank the authors for clarifying my concerns and address my curiosities. The additional results on Resnets at ... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
5,
6,
7,
5
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
4,
3
] | [
"NbtfIuV3ttP",
"l3mv-7mVHH2",
"dHDJMvNsYJj",
"3s37zZZpxdO",
"p3EkXL3cyMG",
"WNbTK4UgSzi",
"A_u3ofv665Z",
"zgEK6m68Fkz",
"1oh6w5xKLfs",
"OKKaDv7xCU6",
"O18hIUDvph8",
"yMd27bY6XN",
"UJMt8Qm8I0R",
"6o35yk8l01T",
"9GpuphXn_ZQ",
"FGa8IibP0Y1",
"nips_2022_tIqzLFf3kk",
"nips_2022_tIqzLFf3... |
nips_2022_k7xZKpYebXL | A Lower Bound of Hash Codes' Performance | As a crucial approach for compact representation learning, hashing has achieved great success in effectiveness and efficiency. Numerous heuristic Hamming space metric learning objectives are designed to obtain high-quality hash codes. Nevertheless, a theoretical analysis of criteria for learning good hash codes remains... | Accept | The paper proposed an interesting lower bound in the learning to hash scenarios and builds on that to show a good algorithm that outperforms several learning to hash methods. There were concerns about the size and scale of experiments which was sufficiently addressed in the rebuttal. The reviewers were not in consensus... | test | [
"BgmspayyLAn",
"3TxPI3xwWY",
"KNtTLynuhZ",
"itw6qpHXtL1",
"mMGVCl4RIEV",
"PoqQ9lOKHZ",
"Wp6w2p96eKX",
"vGD36N2CT52",
"DeDbnTHEMmr",
"maxJY0eNaMZ",
"ZQ7IqAWSO7r",
"jtFHjfFrKz7",
"IdTpMPK4lA_",
"jZQ7bKv66K9",
"8PwhIJl2sn"
] | [
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" We are happy to make this paper better. Actually, we have already polished several sections and improved readability in the rebuttal revision (please refer to `General Response`). Considering on your suggestion, we will go on a further proofreading to check if there needs improvements of the paper.\n\n\\\n\\\nBes... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
7,
3,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
4,
4,
3
] | [
"3TxPI3xwWY",
"KNtTLynuhZ",
"Wp6w2p96eKX",
"8PwhIJl2sn",
"jZQ7bKv66K9",
"IdTpMPK4lA_",
"ZQ7IqAWSO7r",
"8PwhIJl2sn",
"8PwhIJl2sn",
"jZQ7bKv66K9",
"IdTpMPK4lA_",
"nips_2022_k7xZKpYebXL",
"nips_2022_k7xZKpYebXL",
"nips_2022_k7xZKpYebXL",
"nips_2022_k7xZKpYebXL"
] |
nips_2022_yam42JWePu | Fine-Grained Semantically Aligned Vision-Language Pre-Training | Large-scale vision-language pre-training has shown impressive advances in a wide range of downstream tasks. Existing methods mainly model the cross-modal alignment by the similarity of the global representations of images and text, or advanced cross-modal attention upon image and text features. However, they fail to ex... | Accept | This paper addressed the fine-grained visual-language alignment from the perspective of game-theoretic interactions. It received diverse scores with three weak accept and one week reject. The technical novelty is acknowledged by all reviewers. The initial reviews raised concerns about unclear explanations of Shapley, i... | train | [
"pM-U2gFln_c",
"lrsIGPAEu-m",
"_15ypRX5enS",
"9eop7sDY_k",
"LjDZBc55Odz",
"6V1SheG4rmT",
"wELlQD5mr2",
"hiNg_nM4POg",
"xRYba-MmdQu",
"LvLqx2UnYDG",
"EAfUOpmPmNFH",
"AmPztFDZFvN",
"QiQmCgL6x66",
"znyhaAhqVmI",
"GVS7ZjYozdy",
"uv2TveQx4m2",
"9BD34rZhuVo",
"H3J6wa9dBkc",
"BdG819mkmT... | [
"author",
"author",
"author",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"author",
"official_reviewer",
"author",
"author",
"author",
"author",
"author",
"author",
"author",
"author",... | [
" Dear reviewer, since the discussion stage is about to end, do you have any major concerns or suggestions? We are happy to discuss with you.",
" Dear reviewer, do you have any further concerns or suggestions? We are very delighted to discuss with you.",
" We really appreciate your precious time and valuable su... | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
6,
6,
4,
6
] | [
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
-1,
3,
4,
5,
3
] | [
"VZqC7EKJ4wI",
"wELlQD5mr2",
"9eop7sDY_k",
"l5tgVE-tTLW",
"6V1SheG4rmT",
"hiNg_nM4POg",
"VZqC7EKJ4wI",
"xzaspfuARWx",
"AmPztFDZFvN",
"EAfUOpmPmNFH",
"QiQmCgL6x66",
"uv2TveQx4m2",
"znyhaAhqVmI",
"nwondwNlCYw",
"nips_2022_yam42JWePu",
"9BD34rZhuVo",
"H3J6wa9dBkc",
"ygKFprg6Iy5",
"n... |
nips_2022_FurHLDnmC5v | Sample Complexity of Learning Heuristic Functions for Greedy-Best-First and A* Search | Greedy best-first search (GBFS) and A* search (A*) are popular algorithms for path-finding on large graphs. Both use so-called heuristic functions, which estimate how close a vertex is to the goal. While heuristic functions have been handcrafted using domain knowledge, recent studies demonstrate that learning heuristic... | Accept | Strong paper studying the sample complexity of learning heuristic functions for GBFS and A*. The reviewers were especially impressed with the theoretical results and find the paper a worthwhile contribution to this conference. | train | [
"HLCXecSB-eH",
"RB04sS2eTPc",
"ouofRAj0lR3",
"87Oj0VKH2dc",
"16a411ceZGh",
"xF9sijer7F-",
"yqNw-wVhD7",
"pRkxp-LdAdb"
] | [
"official_reviewer",
"official_reviewer",
"author",
"author",
"author",
"official_reviewer",
"official_reviewer",
"official_reviewer"
] | [
" Thank you for your answer. I encourage you to include a version of this in the final version, if accepted.",
" Thanks for your answer, and \nI look forward to reading \"An example of improving the upper bound with special heuristics,\" which may appear in the final version. Initially, I thought this result was ... | [
-1,
-1,
-1,
-1,
-1,
7,
7,
8
] | [
-1,
-1,
-1,
-1,
-1,
1,
3,
2
] | [
"87Oj0VKH2dc",
"16a411ceZGh",
"yqNw-wVhD7",
"pRkxp-LdAdb",
"xF9sijer7F-",
"nips_2022_FurHLDnmC5v",
"nips_2022_FurHLDnmC5v",
"nips_2022_FurHLDnmC5v"
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.