Hugging Face's logo

qualcomm
/
BGNet

Other
PyTorch
deprecated
Model card Files
xet
 Community
qaihm-bot commited on
Commit
ea795e6
·
verified ·
1 Parent(s): 7689704

v0.49.1

Browse files

Deprecation notice.

Files changed (2) hide show
  1. LICENSE +0 -1
  2. README.md +3 -129
LICENSE DELETED
@@ -1 +0,0 @@
1
- The license of the original trained model can be found at https://github.com/thograce/BGNet/blob/master/LICENSE.
 
 
README.md CHANGED
@@ -2,134 +2,8 @@
2
  library_name: pytorch
3
  license: other
4
  tags:
5
- - real_time
6
- - android
7
- pipeline_tag: image-segmentation
8
 
9
  ---
10
-
11
- ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bgnet/web-assets/model_demo.png)
12
-
13
- # BGNet: Optimized for Qualcomm Devices
14
-
15
- BGNet or Boundary-Guided Network, is a model designed for camouflaged object detection. It leverages edge semantics to enhance the representation learning process, making it more effective at identifying objects that blend into their surroundings
16
-
17
- This is based on the implementation of BGNet found [here](https://github.com/thograce/bgnet).
18
- This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/bgnet) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
19
-
20
- Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
21
-
22
- ## Getting Started
23
- Due to licensing restrictions, we cannot distribute pre-exported model assets for this model.
24
- Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/bgnet) Python library to compile and export the model with your own:
25
- - Custom weights (e.g., fine-tuned checkpoints)
26
- - Custom input shapes
27
- - Target device and runtime configurations
28
-
29
- See our repository for [BGNet on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/bgnet) for usage instructions.
30
-
31
-
32
- ## Model Details
33
-
34
- **Model Type:** Model_use_case.semantic_segmentation
35
-
36
- **Model Stats:**
37
- - Model checkpoint: BGNet
38
- - Input resolution: 416x416
39
- - Number of parameters: 77.8M
40
- - Model size (float): 297 MB
41
-
42
- ## Performance Summary
43
- | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
44
- |---|---|---|---|---|---|---
45
- | BGNet | ONNX | float | Snapdragon® X2 Elite | 9.954 ms | 154 - 154 MB | NPU
46
- | BGNet | ONNX | float | Snapdragon® X Elite | 19.496 ms | 153 - 153 MB | NPU
47
- | BGNet | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 14.041 ms | 3 - 323 MB | NPU
48
- | BGNet | ONNX | float | Qualcomm® QCS8550 (Proxy) | 19.222 ms | 0 - 162 MB | NPU
49
- | BGNet | ONNX | float | Qualcomm® QCS9075 | 35.939 ms | 2 - 6 MB | NPU
50
- | BGNet | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 11.398 ms | 3 - 248 MB | NPU
51
- | BGNet | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 9.097 ms | 3 - 255 MB | NPU
52
- | BGNet | ONNX | w8a16 | Snapdragon® X2 Elite | 6.204 ms | 78 - 78 MB | NPU
53
- | BGNet | ONNX | w8a16 | Snapdragon® X Elite | 12.1 ms | 78 - 78 MB | NPU
54
- | BGNet | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 8.513 ms | 0 - 404 MB | NPU
55
- | BGNet | ONNX | w8a16 | Qualcomm® QCS6490 | 2711.244 ms | 337 - 385 MB | CPU
56
- | BGNet | ONNX | w8a16 | Qualcomm® QCS8550 (Proxy) | 11.599 ms | 0 - 83 MB | NPU
57
- | BGNet | ONNX | w8a16 | Qualcomm® QCS9075 | 13.94 ms | 1 - 4 MB | NPU
58
- | BGNet | ONNX | w8a16 | Qualcomm® QCM6690 | 1290.656 ms | 290 - 304 MB | CPU
59
- | BGNet | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 6.925 ms | 0 - 258 MB | NPU
60
- | BGNet | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 1290.658 ms | 246 - 258 MB | CPU
61
- | BGNet | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 4.897 ms | 2 - 271 MB | NPU
62
- | BGNet | ONNX | w8a8 | Snapdragon® X2 Elite | 2.957 ms | 77 - 77 MB | NPU
63
- | BGNet | ONNX | w8a8 | Snapdragon® X Elite | 6.975 ms | 77 - 77 MB | NPU
64
- | BGNet | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 4.593 ms | 0 - 338 MB | NPU
65
- | BGNet | ONNX | w8a8 | Qualcomm® QCS6490 | 442.79 ms | 52 - 150 MB | CPU
66
- | BGNet | ONNX | w8a8 | Qualcomm® QCS8550 (Proxy) | 6.811 ms | 0 - 84 MB | NPU
67
- | BGNet | ONNX | w8a8 | Qualcomm® QCS9075 | 7.242 ms | 0 - 4 MB | NPU
68
- | BGNet | ONNX | w8a8 | Qualcomm® QCM6690 | 384.179 ms | 28 - 44 MB | CPU
69
- | BGNet | ONNX | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 3.869 ms | 0 - 229 MB | NPU
70
- | BGNet | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 336.616 ms | 0 - 14 MB | CPU
71
- | BGNet | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 2.908 ms | 0 - 230 MB | NPU
72
- | BGNet | QNN_DLC | float | Snapdragon® X2 Elite | 10.014 ms | 2 - 2 MB | NPU
73
- | BGNet | QNN_DLC | float | Snapdragon® X Elite | 19.991 ms | 2 - 2 MB | NPU
74
- | BGNet | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 14.183 ms | 0 - 310 MB | NPU
75
- | BGNet | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 115.237 ms | 2 - 238 MB | NPU
76
- | BGNet | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 19.489 ms | 2 - 141 MB | NPU
77
- | BGNet | QNN_DLC | float | Qualcomm® SA8775P | 31.72 ms | 2 - 238 MB | NPU
78
- | BGNet | QNN_DLC | float | Qualcomm® QCS9075 | 37.525 ms | 4 - 8 MB | NPU
79
- | BGNet | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 38.772 ms | 0 - 258 MB | NPU
80
- | BGNet | QNN_DLC | float | Qualcomm® SA7255P | 115.237 ms | 2 - 238 MB | NPU
81
- | BGNet | QNN_DLC | float | Qualcomm® SA8295P | 33.991 ms | 2 - 195 MB | NPU
82
- | BGNet | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 11.09 ms | 2 - 237 MB | NPU
83
- | BGNet | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 8.765 ms | 2 - 247 MB | NPU
84
- | BGNet | QNN_DLC | w8a16 | Snapdragon® X2 Elite | 7.033 ms | 1 - 1 MB | NPU
85
- | BGNet | QNN_DLC | w8a16 | Snapdragon® X Elite | 13.013 ms | 1 - 1 MB | NPU
86
- | BGNet | QNN_DLC | w8a16 | Snapdragon® 8 Gen 3 Mobile | 9.115 ms | 0 - 392 MB | NPU
87
- | BGNet | QNN_DLC | w8a16 | Qualcomm® QCS6490 | 61.566 ms | 3 - 6 MB | NPU
88
- | BGNet | QNN_DLC | w8a16 | Qualcomm® QCS8275 (Proxy) | 36.566 ms | 1 - 258 MB | NPU
89
- | BGNet | QNN_DLC | w8a16 | Qualcomm® QCS8550 (Proxy) | 12.481 ms | 1 - 3 MB | NPU
90
- | BGNet | QNN_DLC | w8a16 | Qualcomm® SA8775P | 12.797 ms | 1 - 259 MB | NPU
91
- | BGNet | QNN_DLC | w8a16 | Qualcomm® QCS9075 | 15.13 ms | 3 - 6 MB | NPU
92
- | BGNet | QNN_DLC | w8a16 | Qualcomm® QCM6690 | 200.797 ms | 1 - 362 MB | NPU
93
- | BGNet | QNN_DLC | w8a16 | Qualcomm® QCS8450 (Proxy) | 20.977 ms | 0 - 375 MB | NPU
94
- | BGNet | QNN_DLC | w8a16 | Qualcomm® SA7255P | 36.566 ms | 1 - 258 MB | NPU
95
- | BGNet | QNN_DLC | w8a16 | Qualcomm® SA8295P | 20.514 ms | 0 - 257 MB | NPU
96
- | BGNet | QNN_DLC | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 6.918 ms | 1 - 249 MB | NPU
97
- | BGNet | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 23.244 ms | 1 - 342 MB | NPU
98
- | BGNet | QNN_DLC | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 5.9 ms | 1 - 258 MB | NPU
99
- | BGNet | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 3.154 ms | 1 - 1 MB | NPU
100
- | BGNet | QNN_DLC | w8a8 | Snapdragon® X Elite | 6.422 ms | 0 - 0 MB | NPU
101
- | BGNet | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 4.361 ms | 0 - 319 MB | NPU
102
- | BGNet | QNN_DLC | w8a8 | Qualcomm® QCS6490 | 26.658 ms | 1 - 3 MB | NPU
103
- | BGNet | QNN_DLC | w8a8 | Qualcomm® QCS8275 (Proxy) | 17.832 ms | 1 - 213 MB | NPU
104
- | BGNet | QNN_DLC | w8a8 | Qualcomm® QCS8550 (Proxy) | 6.174 ms | 1 - 2 MB | NPU
105
- | BGNet | QNN_DLC | w8a8 | Qualcomm® SA8775P | 6.553 ms | 1 - 213 MB | NPU
106
- | BGNet | QNN_DLC | w8a8 | Qualcomm® QCS9075 | 7.332 ms | 0 - 3 MB | NPU
107
- | BGNet | QNN_DLC | w8a8 | Qualcomm® QCM6690 | 116.773 ms | 1 - 291 MB | NPU
108
- | BGNet | QNN_DLC | w8a8 | Qualcomm® QCS8450 (Proxy) | 9.56 ms | 0 - 319 MB | NPU
109
- | BGNet | QNN_DLC | w8a8 | Qualcomm® SA7255P | 17.832 ms | 1 - 213 MB | NPU
110
- | BGNet | QNN_DLC | w8a8 | Qualcomm® SA8295P | 9.988 ms | 0 - 215 MB | NPU
111
- | BGNet | QNN_DLC | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 3.532 ms | 1 - 214 MB | NPU
112
- | BGNet | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 10.468 ms | 1 - 285 MB | NPU
113
- | BGNet | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 2.856 ms | 1 - 214 MB | NPU
114
- | BGNet | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 14.449 ms | 1 - 455 MB | NPU
115
- | BGNet | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 113.907 ms | 1 - 299 MB | NPU
116
- | BGNet | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 19.771 ms | 1 - 3 MB | NPU
117
- | BGNet | TFLITE | float | Qualcomm® SA8775P | 31.971 ms | 1 - 299 MB | NPU
118
- | BGNet | TFLITE | float | Qualcomm® QCS9075 | 34.948 ms | 0 - 159 MB | NPU
119
- | BGNet | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 36.998 ms | 1 - 415 MB | NPU
120
- | BGNet | TFLITE | float | Qualcomm® SA7255P | 113.907 ms | 1 - 299 MB | NPU
121
- | BGNet | TFLITE | float | Qualcomm® SA8295P | 32.757 ms | 1 - 260 MB | NPU
122
- | BGNet | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 11.58 ms | 0 - 300 MB | NPU
123
- | BGNet | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 8.66 ms | 0 - 305 MB | NPU
124
-
125
- ## License
126
- * The license for the original implementation of BGNet can be found
127
- [here](https://github.com/thograce/BGNet/blob/master/LICENSE).
128
-
129
- ## References
130
- * [BGNet: Boundary-Guided Camouflaged Object Detection (IJCAI 2022)](https://arxiv.org/abs/2207.00794)
131
- * [Source Model Implementation](https://github.com/thograce/bgnet)
132
-
133
- ## Community
134
- * Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
135
- * For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).
 
2
  library_name: pytorch
3
  license: other
4
  tags:
5
+ - deprecated
6
+ pipeline_tag: other
 
7
 
8
  ---
9
+ This model is deprecated. Please refer to https://aihub.qualcomm.com for the latest models and updates.