source stringlengths 389 24.7k | summary stringlengths 92 1.61k | source_num_tokens int64 88 5.16k | summary_num_tokens int64 26 445 |
|---|---|---|---|
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | Classical papers in materialized view selection introduce a lattice framework that models and captures dependency (ancestor or descendent) among aggregate views in a multidimensional context @cite_34 @cite_23 @cite_0 @cite_37 . This lattice is greedily browsed with the help of cost models to select the best views to ma... | 1,619 | 162 |
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | A wavelet framework for adaptively representing multidimensional data cubes has also been proposed @cite_40 . This method decomposes data cubes into an indexed hierarchy of wavelet view elements that correspond to partial and residual aggregations of data cubes. An algorithm greedily selects a non-expensive set of wave... | 444 | 196 |
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | Other approaches detect common sub-expressions within workload queries in the relational context @cite_22 @cite_11 @cite_48 . The view selection problem then consists in finding common subexpressions corresponding to intermediary results that are suitable to materialize. However, browsing is very costly and these metho... | 704 | 79 |
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | Finally, the most recent approaches are workload-driven. They syntactically analyze a workload to enumerate relevant candidate views @cite_43 . By exploiting the system's query optimizer, they greedily build a configuration of the most pertinent views. A workload is indeed a good starting point to predict future querie... | 335 | 117 |
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | The index selection problem has been studied for many years in databases @cite_53 @cite_49 @cite_43 @cite_1 @cite_50 @cite_45 @cite_58 . In the more specific context of data warehouses, existing research studies may be clustered into two families: algorithms that optimize maintenance cost @cite_6 and algorithms that op... | 2,150 | 165 |
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | Selecting a set of candidate indexes may be automatic or manual. Warehouse administrators may indeed appeal to their expertise and manually provide, from a given workload, a set of candidate indexes @cite_49 @cite_32 @cite_29 . Such a choice is however subjective. Moreover, the task may be very hard to achieve when the... | 777 | 152 |
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | Ascending greedy methods start from an empty set of candidate indexes @cite_20 @cite_49 @cite_29 @cite_14 . They incrementally add in indexes minimizing cost. This process stops when cost ceases decreasing. Contrarily, descending greedy methods consider the whole set of candidate indexes as a starting point. Then, at e... | 798 | 155 |
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | Genetic algorithms are commonly used to resolve optimization problems. They have been adapted to the index selection problem @cite_45 . The initial population is a set of input indexes (an index is assimilated to an individual). The objective function to optimize is the workload cost corresponding to an index configura... | 1,232 | 158 |
Materialized views and indexes are physical structures for accelerating data access that are casually used in data warehouses. However, these data structures generate some maintenance overhead. They also share the same storage space. Most existing studies about materialized view and index selection consider these struc... | Another approach determines a trade-off between storage space allotted to indexes and materialized views, depending on query definition @cite_48 . According to the authors, the key factors to leverage query optimization is aggregation level, defined by the attribute list of Group by clauses in SQL queries, and the sele... | 351 | 184 |
Collaborative work on unstructured or semi-structured documents, such as in literature corpora or source code, often involves agreed upon templates containing metadata. These templates are not consistent across users and over time. Rule-based parsing of these templates is expensive to maintain and tends to fail as new ... | The algorithmics of finding frequent items or patterns has received much attention. For a survey of the stream-based algorithms, see Cormode and Muthukrishnan [p. 253] corm:whats-hot . Finding frequent patterns robustly is possible using gap constraints @cite_29 . | 445 | 69 |
Wireless sensor networks are often used for environmental monitoring applications. In this context sampling and reconstruction of a physical field is one of the most important problems to solve. We focus on a bandlimited field and find under which conditions on the network topology the reconstruction of the field is su... | Few papers have addressed the problem of sampling and reconstruction in sensor networks. Efficient techniques for spatial sampling in sensor networks are proposed in @cite_2 @cite_4 . In particular @cite_2 presents an algorithm to determine which sensor subsets should be selected to acquire data from an area of interes... | 978 | 267 |
Wireless sensor networks are often used for environmental monitoring applications. In this context sampling and reconstruction of a physical field is one of the most important problems to solve. We focus on a bandlimited field and find under which conditions on the network topology the reconstruction of the field is su... | Note that our work significantly differs from the studies above because we assume that the sensors location are known (or can be determined @cite_11 @cite_6 @cite_0 ) and the sensor precision is sufficiently high so that the quantization error is negligible. The question we pose is instead under which conditions (on th... | 547 | 89 |
As the Internet struggles to cope with scalability, mobility, and security issues, new network architectures are being proposed to better accommodate the needs of modern systems and applications. In particular, Content-Oriented Networking (CON) has emerged as a promising next-generation Internet architecture: it sets t... | Propelled by the increasing interest for next-generation Internet architectures and, in particular, Content-Oriented Networking (CON), the research community has produced a large body of work dealing with CON building blocks @cite_27 @cite_3 @cite_33 @cite_44 @cite_24 , performance @cite_78 @cite_49 @cite_72 @cite_82 ,... | 1,849 | 144 |
As the Internet struggles to cope with scalability, mobility, and security issues, new network architectures are being proposed to better accommodate the needs of modern systems and applications. In particular, Content-Oriented Networking (CON) has emerged as a promising next-generation Internet architecture: it sets t... | Security in CON. Wong and Nikander @cite_25 address security of naming mechanisms by constructing content name as the concatenation of content provider's ID, cryptographic ID of the content and some meta-data. @cite_5 adopt a similar approach where content name is defined as the concatenation of the hash of the public ... | 1,169 | 272 |
As the Internet struggles to cope with scalability, mobility, and security issues, new network architectures are being proposed to better accommodate the needs of modern systems and applications. In particular, Content-Oriented Networking (CON) has emerged as a promising next-generation Internet architecture: it sets t... | Privacy Issues in CON. To the best of our knowledge, the only related privacy study is the recent article by in @cite_48 @cite_37 , that covers security and privacy issues of CCN @cite_27 . Specifically, they highlight a few Denial-of-Service (DoS) vulnerabilities as well as different cache-related attacks. In CCN, a p... | 530 | 228 |
As the Internet struggles to cope with scalability, mobility, and security issues, new network architectures are being proposed to better accommodate the needs of modern systems and applications. In particular, Content-Oriented Networking (CON) has emerged as a promising next-generation Internet architecture: it sets t... | From the privacy perspective, work in @cite_48 @cite_37 identifies the issue of information leakage through caches in CCN. It proposes a few simple countermeasures, following and approaches. The former can be achieved using techniques similar to those addressing cache pollution attacks in IP @cite_53 , although such an... | 668 | 280 |
As the Internet struggles to cope with scalability, mobility, and security issues, new network architectures are being proposed to better accommodate the needs of modern systems and applications. In particular, Content-Oriented Networking (CON) has emerged as a promising next-generation Internet architecture: it sets t... | Our work extends that in @cite_48 @cite_37 by encompassing all privacy aspects: caching, naming, signature, and content. Also, it is more general as it does not only consider CCN @cite_27 , but CON in general, independently of the specific instantiation. Furthermore, when suggesting countermeasures, we only propose tec... | 530 | 91 |
As the Internet struggles to cope with scalability, mobility, and security issues, new network architectures are being proposed to better accommodate the needs of modern systems and applications. In particular, Content-Oriented Networking (CON) has emerged as a promising next-generation Internet architecture: it sets t... | Anonymity in CON. AND @math NA @cite_30 proposes a Tor-like anonymizing tool for CCN @cite_27 to provide provable anonymity. It also aims to privacy protection via simple tunneling. However, as discussed in , AND @math NA is an all-in-one'' solution that introduces latency and impedes caching. Whereas, fine-grained pri... | 718 | 186 |
A smooth affine hypersurface Z of complex dimension n is homotopy equivalent to an n-dimensional cell complex. Given a defining polynomial f for Z as well as a regular triangulation of its Newton polytope, we provide a purely combinatorial construction of a compact topological space S as a union of components of real d... | A skeleton for Fermat hypersurfaces was described by Deligne in [pp. 88--90] Deligne , and this skeleton is visible in our own in a manner described in Remark . Our skeleta'' are different than the skeleta'' that appear in nonarchimedean geometry @cite_10 @cite_15 , but @math plays a similar role in both constructions.... | 408 | 114 |
A smooth affine hypersurface Z of complex dimension n is homotopy equivalent to an n-dimensional cell complex. Given a defining polynomial f for Z as well as a regular triangulation of its Newton polytope, we provide a purely combinatorial construction of a compact topological space S as a union of components of real d... | Hypersurfaces in algebraic tori have been studied by Danilov-Khovanski @cite_14 and Batyrev @cite_19 . Danilov-Khovankski computed mixed Hodge numbers, while Batyev studied the variation of mixed Hodge structures. Log geometry has been extensively employed by Gross and Siebert @cite_21 in their seminal work studying th... | 933 | 271 |
A smooth affine hypersurface Z of complex dimension n is homotopy equivalent to an n-dimensional cell complex. Given a defining polynomial f for Z as well as a regular triangulation of its Newton polytope, we provide a purely combinatorial construction of a compact topological space S as a union of components of real d... | In the symplectic-topological setting, Mikhalkin @cite_5 constructed a degeneration of a projective algebraic hypersurface using a triangulation of its Newton polytope to provide a higher-dimensional pair-of-pants'' decomposition. He further identified a stratified torus fibration over the spine of the corresponding am... | 442 | 154 |
Many online social networks feature restrictive web interfaces which only allow the query of a user's local neighborhood through the interface. To enable analytics over such an online social network through its restrictive web interface, many recent efforts reuse the existing Markov Chain Monte Carlo methods such as ra... | With global topology, @cite_12 discussed sampling techniques like random node, random edge, random subgraph in large graphs. @cite_22 introduced Albatross sampling which combines random jump and MHRW. @cite_26 also demonstrated true uniform sampling method among the users' id as ground-truth". | 846 | 72 |
Many online social networks feature restrictive web interfaces which only allow the query of a user's local neighborhood through the interface. To enable analytics over such an online social network through its restrictive web interface, many recent efforts reuse the existing Markov Chain Monte Carlo methods such as ra... | @cite_21 found that the mixing time of typical online social networks is much larger than anticipated, which validates our motivation to shorten the mixing time of random walk. @cite_13 derived the fastest mixing random walk on a graph by convex optimization on second largest eigenvalue of the transition matrix, but it... | 933 | 95 |
Many online social networks feature restrictive web interfaces which only allow the query of a user's local neighborhood through the interface. To enable analytics over such an online social network through its restrictive web interface, many recent efforts reuse the existing Markov Chain Monte Carlo methods such as ra... | @cite_6 compared latent space model with real social network data. @cite_14 introduced hybrid graph model to incorporate the small world phenomenon. @cite_7 also measured the difference between multiple synthetic graphs and real world social network graphs. | 827 | 53 |
We consider the problem of recovering two unknown vectors, @math and @math , of length @math from their circular convolution. We make the structural assumption that the two vectors are members of known subspaces, one with dimension @math and the other with dimension @math . Although the observed convolution is nonlinea... | While this paper is only concerned with recovery by nuclear norm minimization, other types of recovery techniques have proven effective both in theory and in practice; see for example @cite_4 @cite_12 @cite_14 . It is possible that the guarantees given in this paper could be extended to these other algorithms. | 1,213 | 65 |
We consider the problem of recovering two unknown vectors, @math and @math , of length @math from their circular convolution. We make the structural assumption that the two vectors are members of known subspaces, one with dimension @math and the other with dimension @math . Although the observed convolution is nonlinea... | As we will see below, our mathematical analysis has mostly to do how matrices of the form in act on rank-2 matrices in a certain subspace. Matrices of this type have been considered in the context of sparse recovery in the compressed sensing literature for applications including multiple-input multiple-output channel e... | 855 | 109 |
We introduce a new family of erasure codes, called group decodable code (GDC), for distributed storage system. Given a set of design parameters ; ; k; t , where k is the number of information symbols, each codeword of an ( ; ; k; t)-group decodable code is a t-tuple of strings, called buckets, such that each bucket is ... | In @cite_12 , the concept of @math -locality was defined, which captures the property that there exist @math pairwise disjoint local repair sets for a code symbol. An upper bound on the minimum distance for @math linear codes with information @math -locality was derived, and codes that attain this bound was constructed... | 688 | 176 |
This paper aims at developing an integrated system of clothing co-parsing, in order to jointly parse a set of clothing images (unsegmented but annotated with tags) into semantic configurations. We propose a data-driven framework consisting of two phases of inference. The first phase, referred as "image co-segmentation"... | In literature, existing efforts on clothing human segmentation and recognition mainly focused on constructing expressive models to address various clothing styles and appearances @cite_0 @cite_1 @cite_14 @cite_34 @cite_32 @cite_3 @cite_19 . One classic work @cite_0 proposed a composite And-Or graph template for modelin... | 1,203 | 196 |
This paper aims at developing an integrated system of clothing co-parsing, in order to jointly parse a set of clothing images (unsegmented but annotated with tags) into semantic configurations. We propose a data-driven framework consisting of two phases of inference. The first phase, referred as "image co-segmentation"... | Clothing co-parsing is also highly related to image object co-labeling, where a batch of input images containing similar objects are processed jointly @cite_5 @cite_21 @cite_27 . For example, unsupervised shape guided approaches were adopted in @cite_13 to achieve single object category co-labeling. Winn et. al. @cite_... | 1,991 | 268 |
This article investigates a data-driven approach for semantic scene understanding, without pixelwise annotation or classifier training. The proposed framework parses a target image in two steps: first, retrieving its exemplars (that is, references) from an image database, where all images are unsegmented but annotated ... | Traditional efforts for scene understanding mainly focused on capturing scene appearances, structures and spatial contexts by developing combinatorial models, e.g., CRF @cite_9 @cite_17 , Texton-Forest @cite_0 , Graph Grammar @cite_4 . These models were generally founded on supervised learning techniques, and required ... | 1,145 | 91 |
This article investigates a data-driven approach for semantic scene understanding, without pixelwise annotation or classifier training. The proposed framework parses a target image in two steps: first, retrieving its exemplars (that is, references) from an image database, where all images are unsegmented but annotated ... | Several weakly supervised methods are proposed to indicate the classes presented in the images with only image-level labels. For example, @cite_16 proposed to learn object classes based on unsupervised image segmentation. @cite_8 learned classification models for all scene labels by selecting representative training sa... | 1,153 | 143 |
This paper investigates how to extract objects-of-interest without relying on hand-craft features and sliding windows approaches, that aims to jointly solve two sub-tasks: (i) rapidly localizing salient objects from images, and (ii) accurately segmenting the objects based on the localizations. We present a general join... | Extracting pixelwise objects-of-interest from an image, our work is related to the salient region object detections @cite_34 @cite_19 @cite_0 @cite_26 . These methods mainly focused on feature engineering and graph-based segmentation. For example, @cite_19 proposed a regional contrast based saliency extraction algorith... | 1,891 | 192 |
This paper investigates how to extract objects-of-interest without relying on hand-craft features and sliding windows approaches, that aims to jointly solve two sub-tasks: (i) rapidly localizing salient objects from images, and (ii) accurately segmenting the objects based on the localizations. We present a general join... | Recently resurgent deep learning methods have also been applied in object detection and image segmentation @cite_30 @cite_25 @cite_15 @cite_13 @cite_18 @cite_3 @cite_27 @cite_2 . Among these works, @cite_8 detected objects by training category-level convolutional neural networks. @cite_2 proposed to combine multiple co... | 1,576 | 233 |
In this paper, we propose a novel model for high-dimensional data, called the Hybrid Orthogonal Projection and Estimation (HOPE) model, which combines a linear orthogonal projection and a finite mixture model under a unified generative modeling framework. The HOPE model itself can be learned unsupervised from unlabelle... | Similar to PCA, the Fisher's linear discriminant analysis (LDA) can also be viewed as a linear dimensionality reduction technique. However, PCA is unsupervised in the sense that PCA depends only on the data while Fisher's LDA is supervised since it uses both data and class-label information. The high-dimensional data a... | 435 | 172 |
The problem of influence maximization is to select the most influential individuals in a social network. With the popularity of social network sites, and the development of viral marketing, the importance of the problem has been increased. The influence maximization problem is NP-hard, and therefore, there will not exi... | In order to improve the efficiency of the computations, many studies have been made. Leskovec al @cite_3 proposed Cost-Effective Lazy Forward (CELF) optimization that reduces the computation cost of the influence spread using sub-modularity property of the objective function. | 533 | 63 |
The problem of influence maximization is to select the most influential individuals in a social network. With the popularity of social network sites, and the development of viral marketing, the importance of the problem has been increased. The influence maximization problem is NP-hard, and therefore, there will not exi... | Chen al @cite_8 proposed new greedy algorithms for independent cascade and weighted cascade models. They made the greedy algorithm faster by combining their algorithms with CELF. They also proposed a new heuristic, named , which produces results of quality close to the greedy algorithm, while is much faster than that a... | 2,487 | 280 |
The problem of influence maximization is to select the most influential individuals in a social network. With the popularity of social network sites, and the development of viral marketing, the importance of the problem has been increased. The influence maximization problem is NP-hard, and therefore, there will not exi... | There are some community-based algorithms for the influence maximization problem that partition the graph into small subgraphs and select the most influential vertices from each subgraph. Chen al @cite_2 used H-Clustering algorithm and Manaskasemsak al @cite_13 used Markov clustering algorithm for community detection. ... | 1,046 | 102 |
We consider the problem of learning an unknown function @math on the @math -dimensional sphere with respect to the square loss, given i.i.d. samples @math where @math is a feature vector uniformly distributed on the sphere and @math . We study two popular classes of models that can be regarded as linearizations of two-... | Approximation properties of neural networks and, more generally, nonlinear approximation have been studied in detail in the nineties, see e.g. @cite_17 @cite_2 @cite_4 . The main concern of the present paper is quite different, since we focus on the random feature model, and the (recently proposed) neural tangent model... | 985 | 128 |
We consider the problem of learning an unknown function @math on the @math -dimensional sphere with respect to the square loss, given i.i.d. samples @math where @math is a feature vector uniformly distributed on the sphere and @math . We study two popular classes of models that can be regarded as linearizations of two-... | The random features model , has been studied in considerable depth since the original work in @cite_21 . The classical viewpoint suggests that @math should be regarded as an approximation of the reproducing kernel Hilbert space (RKHS) @math defined by the kernel (see @cite_13 for general background) Indeed the space @m... | 1,488 | 200 |
We consider the problem of learning an unknown function @math on the @math -dimensional sphere with respect to the square loss, given i.i.d. samples @math where @math is a feature vector uniformly distributed on the sphere and @math . We study two popular classes of models that can be regarded as linearizations of two-... | Of course, this approach generally breaks down if the dimension @math is large (technically, if it grows with @math ). This curse of dimensionality' is already revealed by classical lower bounds in functional approximation, see e.g. @cite_17 @cite_18 . However, previous work does not clarify what happens precisely in t... | 683 | 128 |
We consider the problem of learning an unknown function @math on the @math -dimensional sphere with respect to the square loss, given i.i.d. samples @math where @math is a feature vector uniformly distributed on the sphere and @math . We study two popular classes of models that can be regarded as linearizations of two-... | The connection between kernel methods and neural networks was recently revived by the work of Belkin and coauthors @cite_12 @cite_24 who pointed out intriguing similarities between some properties of modern deep learning models, and large scale kernel learning. A concrete explanation for this analogy was proposed in @c... | 1,888 | 260 |
We consider the problem of learning an unknown function @math on the @math -dimensional sphere with respect to the square loss, given i.i.d. samples @math where @math is a feature vector uniformly distributed on the sphere and @math . We study two popular classes of models that can be regarded as linearizations of two-... | It is worth mentioning that an alternative approach to the analysis of two-layers neural networks, in the limit of a large number of neurons, was developed in @cite_19 @cite_14 @cite_20 @cite_22 @cite_9 . Unlike in the neural tangent approach, the evolution of network weights is described beyond the linear regime in th... | 1,401 | 83 |
The abundance of open-source code, coupled with the success of recent advances in deep learning for natural language processing, has given rise to a promising new application of machine learning to source code. In this work, we explore the use of a Siamese recurrent neural network model on Python source code to create ... | introduce CODEnn @cite_7 , which creates vector representations of Java source code by jointly embedding code with a natural language description of the method. Their architecture uses recurrent neural networks (RNN) on sequences of API calls and on tokens from the method name. It then fuses this with the output of a m... | 466 | 193 |
The abundance of open-source code, coupled with the success of recent advances in deep learning for natural language processing, has given rise to a promising new application of machine learning to source code. In this work, we explore the use of a Siamese recurrent neural network model on Python source code to create ... | @cite_0 convert the AST into a binary tree and then use an autoencoder to learn an embedding model for each node. This learned embedding model is applied recursively to the tree to obtain a final embedding for the root node. The model is an autoencoder, and as such, may fail to recognize the semantic equivalency betwee... | 425 | 116 |
The abundance of open-source code, coupled with the success of recent advances in deep learning for natural language processing, has given rise to a promising new application of machine learning to source code. In this work, we explore the use of a Siamese recurrent neural network model on Python source code to create ... | The work of @cite_4 encodes nodes of an AST using a weighted mixture of left and right weight matrices and children of that node. A tree-based convolutional neural network (CNN) is then applied against the tree to encode the AST. The only way semantic equivalents are learned in this model are by recognizing that certai... | 401 | 109 |
The abundance of open-source code, coupled with the success of recent advances in deep learning for natural language processing, has given rise to a promising new application of machine learning to source code. In this work, we explore the use of a Siamese recurrent neural network model on Python source code to create ... | @cite_1 used a long-short term memory (LSTM) cell in a tree structure applied to an AST to classify defects. The model is trained in an unsupervised manner to predict a node from its children. @cite_13 learn code vector embeddings by evaluating paths in the AST and evaluate the resulting vectors by predicting method na... | 905 | 142 |
The abundance of open-source code, coupled with the success of recent advances in deep learning for natural language processing, has given rise to a promising new application of machine learning to source code. In this work, we explore the use of a Siamese recurrent neural network model on Python source code to create ... | In Deep Code Comment Generation @cite_11 , introduce structure-based traversals of an AST to feed into a sequence to sequence architecture, and train the model to translate source code into comments. Then the trained model is used to generate comments on new source code. The encoded vectors are designed to initialize a... | 340 | 110 |
As an intuitive way of expression emotion, the animated Graphical Interchange Format (GIF) images have been widely used on social media. Most previous studies on automated GIF emotion recognition fail to effectively utilize GIF's unique properties, and this potentially limits the recognition performance. In this study,... | GIF Analysis. Bakhshi @cite_13 show that animated GIFs are more engaging than other social media types by studying over 3.9 million posts on Tumblr. Gygli @cite_12 propose to automatically generate animated GIFs from videos with 100K user-generated GIFs and the corresponding video sources. The MIT's GIFGIF platform is ... | 1,300 | 176 |
As an intuitive way of expression emotion, the animated Graphical Interchange Format (GIF) images have been widely used on social media. Most previous studies on automated GIF emotion recognition fail to effectively utilize GIF's unique properties, and this potentially limits the recognition performance. In this study,... | Emotion Recognition. Emotion recognition @cite_0 @cite_6 has been an interesting topic for decades. On a large scale dataset @cite_23 , Rao @cite_8 propose a multi-level deep representations for emotion recognition. Multi-modal feature fusion @cite_22 is also proved to be effective. Instead of modeling emotion recognit... | 2,008 | 160 |
Magnetic resonance imaging (MRI) is being increasingly utilized to assess, diagnose, and plan treatment for a variety of diseases. The ability to visualize tissue in varied contrasts in the form of MR pulse sequences in a single scan provides valuable insights to physicians, as well as enabling automated systems perfor... | Though all the methods discussed above propose a multi-input method, none of the methods have been proposed to synthesize multiple missing sequences (multi-output), and in one single pass. All three methods @cite_50 , @cite_52 , and @cite_2 synthesize only sequence in the presence of varying number of input sequences, ... | 1,432 | 203 |
Magnetic resonance imaging (MRI) is being increasingly utilized to assess, diagnose, and plan treatment for a variety of diseases. The ability to visualize tissue in varied contrasts in the form of MR pulse sequences in a single scan provides valuable insights to physicians, as well as enabling automated systems perfor... | The main motivation for most synthesis methods is to retain the ability to meaningfully use some downstream analysis pipelines like segmentation or classification despite the partially missing input. However, there have been efforts by researchers working on those analysis pipelines to bypass any synthesis step by maki... | 722 | 234 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.