Hugging Face's logo

qualcomm
/
DETR-ResNet101

Object Detection
PyTorch
bu_auto
android
Model card Files
xet
 Community
qaihm-bot commited on
Commit
0fa16c2
·
verified ·
1 Parent(s): 9bb3f05

v0.48.0

Browse files

See https://github.com/qualcomm/ai-hub-models/releases/v0.48.0 for changelog.

Files changed (1) hide show
  1. README.md +35 -35
README.md CHANGED
@@ -15,7 +15,7 @@ pipeline_tag: object-detection
15
  DETR is a machine learning model that can detect objects (trained on COCO dataset).
16
 
17
  This is based on the implementation of DETR-ResNet101 found [here](https://github.com/facebookresearch/detr).
18
- This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/detr_resnet101) 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
 
@@ -28,23 +28,23 @@ Below are pre-exported model assets ready for deployment.
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
- | ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/detr_resnet101/releases/v0.47.0/detr_resnet101-onnx-float.zip)
32
- | QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/detr_resnet101/releases/v0.47.0/detr_resnet101-qnn_dlc-float.zip)
33
- | TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/detr_resnet101/releases/v0.47.0/detr_resnet101-tflite-float.zip)
34
 
35
  For more device-specific assets and performance metrics, visit **[DETR-ResNet101 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/detr_resnet101)**.
36
 
37
 
38
  ### Option 2: Export with Custom Configurations
39
 
40
- Use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/detr_resnet101) Python library to compile and export the model with your own:
41
  - Custom weights (e.g., fine-tuned checkpoints)
42
  - Custom input shapes
43
  - Target device and runtime configurations
44
 
45
  This option is ideal if you need to customize the model beyond the default configuration provided here.
46
 
47
- See our repository for [DETR-ResNet101 on GitHub](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/detr_resnet101) for usage instructions.
48
 
49
  ## Model Details
50
 
@@ -59,35 +59,35 @@ See our repository for [DETR-ResNet101 on GitHub](https://github.com/quic/ai-hub
59
  ## Performance Summary
60
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
61
  |---|---|---|---|---|---|---
62
- | DETR-ResNet101 | ONNX | float | Snapdragon® X Elite | 24.857 ms | 114 - 114 MB | NPU
63
- | DETR-ResNet101 | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 18.645 ms | 1 - 523 MB | NPU
64
- | DETR-ResNet101 | ONNX | float | Qualcomm® QCS8550 (Proxy) | 24.744 ms | 5 - 10 MB | NPU
65
- | DETR-ResNet101 | ONNX | float | Qualcomm® QCS9075 | 39.664 ms | 5 - 12 MB | NPU
66
- | DETR-ResNet101 | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 14.662 ms | 0 - 357 MB | NPU
67
- | DETR-ResNet101 | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 11.801 ms | 5 - 359 MB | NPU
68
- | DETR-ResNet101 | ONNX | float | Snapdragon® X2 Elite | 12.832 ms | 115 - 115 MB | NPU
69
- | DETR-ResNet101 | QNN_DLC | float | Snapdragon® X Elite | 27.183 ms | 5 - 5 MB | NPU
70
- | DETR-ResNet101 | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 19.902 ms | 1 - 466 MB | NPU
71
- | DETR-ResNet101 | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 138.112 ms | 1 - 336 MB | NPU
72
- | DETR-ResNet101 | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 26.98 ms | 5 - 191 MB | NPU
73
- | DETR-ResNet101 | QNN_DLC | float | Qualcomm® SA8775P | 41.291 ms | 1 - 336 MB | NPU
74
- | DETR-ResNet101 | QNN_DLC | float | Qualcomm® QCS9075 | 46.011 ms | 5 - 11 MB | NPU
75
- | DETR-ResNet101 | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 56.597 ms | 2 - 366 MB | NPU
76
- | DETR-ResNet101 | QNN_DLC | float | Qualcomm® SA7255P | 138.112 ms | 1 - 336 MB | NPU
77
- | DETR-ResNet101 | QNN_DLC | float | Qualcomm® SA8295P | 43.601 ms | 0 - 251 MB | NPU
78
- | DETR-ResNet101 | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 14.985 ms | 5 - 344 MB | NPU
79
- | DETR-ResNet101 | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 11.786 ms | 5 - 358 MB | NPU
80
- | DETR-ResNet101 | QNN_DLC | float | Snapdragon® X2 Elite | 13.382 ms | 5 - 5 MB | NPU
81
- | DETR-ResNet101 | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 19.717 ms | 0 - 528 MB | NPU
82
- | DETR-ResNet101 | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 137.851 ms | 0 - 396 MB | NPU
83
- | DETR-ResNet101 | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 26.948 ms | 0 - 3 MB | NPU
84
- | DETR-ResNet101 | TFLITE | float | Qualcomm® SA8775P | 41.163 ms | 0 - 395 MB | NPU
85
- | DETR-ResNet101 | TFLITE | float | Qualcomm® QCS9075 | 41.608 ms | 0 - 125 MB | NPU
86
- | DETR-ResNet101 | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 56.108 ms | 3 - 420 MB | NPU
87
- | DETR-ResNet101 | TFLITE | float | Qualcomm® SA7255P | 137.851 ms | 0 - 396 MB | NPU
88
- | DETR-ResNet101 | TFLITE | float | Qualcomm® SA8295P | 43.414 ms | 0 - 307 MB | NPU
89
- | DETR-ResNet101 | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 15.184 ms | 0 - 404 MB | NPU
90
- | DETR-ResNet101 | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 11.264 ms | 0 - 416 MB | NPU
91
 
92
  ## License
93
  * The license for the original implementation of DETR-ResNet101 can be found
 
15
  DETR is a machine learning model that can detect objects (trained on COCO dataset).
16
 
17
  This is based on the implementation of DETR-ResNet101 found [here](https://github.com/facebookresearch/detr).
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/detr_resnet101) 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
 
 
28
 
29
  | Runtime | Precision | Chipset | SDK Versions | Download |
30
  |---|---|---|---|---|
31
+ | ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/detr_resnet101/releases/v0.48.0/detr_resnet101-onnx-float.zip)
32
+ | QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/detr_resnet101/releases/v0.48.0/detr_resnet101-qnn_dlc-float.zip)
33
+ | TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/detr_resnet101/releases/v0.48.0/detr_resnet101-tflite-float.zip)
34
 
35
  For more device-specific assets and performance metrics, visit **[DETR-ResNet101 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/detr_resnet101)**.
36
 
37
 
38
  ### Option 2: Export with Custom Configurations
39
 
40
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/detr_resnet101) Python library to compile and export the model with your own:
41
  - Custom weights (e.g., fine-tuned checkpoints)
42
  - Custom input shapes
43
  - Target device and runtime configurations
44
 
45
  This option is ideal if you need to customize the model beyond the default configuration provided here.
46
 
47
+ See our repository for [DETR-ResNet101 on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/detr_resnet101) for usage instructions.
48
 
49
  ## Model Details
50
 
 
59
  ## Performance Summary
60
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
61
  |---|---|---|---|---|---|---
62
+ | DETR-ResNet101 | ONNX | float | Snapdragon® X2 Elite | 12.797 ms | 115 - 115 MB | NPU
63
+ | DETR-ResNet101 | ONNX | float | Snapdragon® X Elite | 24.913 ms | 114 - 114 MB | NPU
64
+ | DETR-ResNet101 | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 18.725 ms | 1 - 523 MB | NPU
65
+ | DETR-ResNet101 | ONNX | float | Qualcomm® QCS8550 (Proxy) | 24.649 ms | 0 - 127 MB | NPU
66
+ | DETR-ResNet101 | ONNX | float | Qualcomm® QCS9075 | 40.476 ms | 5 - 12 MB | NPU
67
+ | DETR-ResNet101 | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 14.451 ms | 0 - 356 MB | NPU
68
+ | DETR-ResNet101 | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 11.769 ms | 5 - 359 MB | NPU
69
+ | DETR-ResNet101 | QNN_DLC | float | Snapdragon® X2 Elite | 13.383 ms | 5 - 5 MB | NPU
70
+ | DETR-ResNet101 | QNN_DLC | float | Snapdragon® X Elite | 27.158 ms | 5 - 5 MB | NPU
71
+ | DETR-ResNet101 | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 19.904 ms | 5 - 474 MB | NPU
72
+ | DETR-ResNet101 | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 138.114 ms | 0 - 337 MB | NPU
73
+ | DETR-ResNet101 | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 26.855 ms | 5 - 448 MB | NPU
74
+ | DETR-ResNet101 | QNN_DLC | float | Qualcomm® SA8775P | 41.253 ms | 1 - 337 MB | NPU
75
+ | DETR-ResNet101 | QNN_DLC | float | Qualcomm® QCS9075 | 42.33 ms | 5 - 11 MB | NPU
76
+ | DETR-ResNet101 | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 56.687 ms | 3 - 365 MB | NPU
77
+ | DETR-ResNet101 | QNN_DLC | float | Qualcomm® SA7255P | 138.114 ms | 0 - 337 MB | NPU
78
+ | DETR-ResNet101 | QNN_DLC | float | Qualcomm® SA8295P | 43.546 ms | 0 - 251 MB | NPU
79
+ | DETR-ResNet101 | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 15.077 ms | 5 - 346 MB | NPU
80
+ | DETR-ResNet101 | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 11.792 ms | 4 - 360 MB | NPU
81
+ | DETR-ResNet101 | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 19.946 ms | 0 - 533 MB | NPU
82
+ | DETR-ResNet101 | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 137.886 ms | 0 - 397 MB | NPU
83
+ | DETR-ResNet101 | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 26.675 ms | 0 - 3 MB | NPU
84
+ | DETR-ResNet101 | TFLITE | float | Qualcomm® SA8775P | 41.096 ms | 0 - 397 MB | NPU
85
+ | DETR-ResNet101 | TFLITE | float | Qualcomm® QCS9075 | 41.615 ms | 0 - 125 MB | NPU
86
+ | DETR-ResNet101 | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 55.105 ms | 0 - 419 MB | NPU
87
+ | DETR-ResNet101 | TFLITE | float | Qualcomm® SA7255P | 137.886 ms | 0 - 397 MB | NPU
88
+ | DETR-ResNet101 | TFLITE | float | Qualcomm® SA8295P | 43.399 ms | 0 - 307 MB | NPU
89
+ | DETR-ResNet101 | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 15.187 ms | 0 - 409 MB | NPU
90
+ | DETR-ResNet101 | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 11.307 ms | 0 - 415 MB | NPU
91
 
92
  ## License
93
  * The license for the original implementation of DETR-ResNet101 can be found