Hugging Face's logo

qualcomm
/
CavaFace

Object Detection
PyTorch
real_time
bu_auto
android
Model card Files
xet
 Community
qaihm-bot commited on
Commit
afe0522
·
verified ·
1 Parent(s): 26229f5

v0.49.1

Browse files

See https://github.com/qualcomm/ai-hub-models/releases/v0.49.1 for changelog.

Files changed (1) hide show
  1. README.md +35 -35
README.md CHANGED
@@ -16,7 +16,7 @@ pipeline_tag: object-detection
16
  A PyTorch-based framework for training face recognition models that generates facial embeddings for verification and identification tasks
17
 
18
  This is based on the implementation of CavaFace found [here](https://github.com/cavalleria/cavaface).
19
- 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/cavaface) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
20
 
21
  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.
22
 
@@ -29,23 +29,23 @@ Below are pre-exported model assets ready for deployment.
29
 
30
  | Runtime | Precision | Chipset | SDK Versions | Download |
31
  |---|---|---|---|---|
32
- | 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/cavaface/releases/v0.48.0/cavaface-onnx-float.zip)
33
- | QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/cavaface/releases/v0.48.0/cavaface-qnn_dlc-float.zip)
34
- | 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/cavaface/releases/v0.48.0/cavaface-tflite-float.zip)
35
 
36
  For more device-specific assets and performance metrics, visit **[CavaFace on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/cavaface)**.
37
 
38
 
39
  ### Option 2: Export with Custom Configurations
40
 
41
- Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/cavaface) Python library to compile and export the model with your own:
42
  - Custom weights (e.g., fine-tuned checkpoints)
43
  - Custom input shapes
44
  - Target device and runtime configurations
45
 
46
  This option is ideal if you need to customize the model beyond the default configuration provided here.
47
 
48
- See our repository for [CavaFace on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/cavaface) for usage instructions.
49
 
50
  ## Model Details
51
 
@@ -60,35 +60,35 @@ See our repository for [CavaFace on GitHub](https://github.com/qualcomm/ai-hub-m
60
  ## Performance Summary
61
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
62
  |---|---|---|---|---|---|---
63
- | CavaFace | ONNX | float | Snapdragon® X2 Elite | 2.337 ms | 126 - 126 MB | NPU
64
- | CavaFace | ONNX | float | Snapdragon® X Elite | 4.511 ms | 126 - 126 MB | NPU
65
- | CavaFace | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 3.2 ms | 0 - 110 MB | NPU
66
- | CavaFace | ONNX | float | Qualcomm® QCS8550 (Proxy) | 4.34 ms | 0 - 131 MB | NPU
67
- | CavaFace | ONNX | float | Qualcomm® QCS9075 | 6.789 ms | 0 - 3 MB | NPU
68
- | CavaFace | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.64 ms | 0 - 79 MB | NPU
69
- | CavaFace | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.265 ms | 0 - 92 MB | NPU
70
- | CavaFace | QNN_DLC | float | Snapdragon® X2 Elite | 2.602 ms | 0 - 0 MB | NPU
71
- | CavaFace | QNN_DLC | float | Snapdragon® X Elite | 4.456 ms | 0 - 0 MB | NPU
72
- | CavaFace | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 3.187 ms | 0 - 102 MB | NPU
73
- | CavaFace | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 24.701 ms | 0 - 81 MB | NPU
74
- | CavaFace | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 4.299 ms | 0 - 2 MB | NPU
75
- | CavaFace | QNN_DLC | float | Qualcomm® SA8775P | 30.193 ms | 0 - 81 MB | NPU
76
- | CavaFace | QNN_DLC | float | Qualcomm® QCS9075 | 6.772 ms | 0 - 2 MB | NPU
77
- | CavaFace | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 8.941 ms | 0 - 109 MB | NPU
78
- | CavaFace | QNN_DLC | float | Qualcomm® SA7255P | 24.701 ms | 0 - 81 MB | NPU
79
- | CavaFace | QNN_DLC | float | Qualcomm® SA8295P | 7.956 ms | 0 - 86 MB | NPU
80
- | CavaFace | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.63 ms | 0 - 85 MB | NPU
81
- | CavaFace | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.235 ms | 0 - 84 MB | NPU
82
- | CavaFace | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 3.143 ms | 0 - 213 MB | NPU
83
- | CavaFace | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 24.552 ms | 0 - 98 MB | NPU
84
- | CavaFace | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 4.172 ms | 0 - 2 MB | NPU
85
- | CavaFace | TFLITE | float | Qualcomm® SA8775P | 6.896 ms | 0 - 98 MB | NPU
86
- | CavaFace | TFLITE | float | Qualcomm® QCS9075 | 6.699 ms | 0 - 129 MB | NPU
87
- | CavaFace | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 8.752 ms | 0 - 221 MB | NPU
88
- | CavaFace | TFLITE | float | Qualcomm® SA7255P | 24.552 ms | 0 - 98 MB | NPU
89
- | CavaFace | TFLITE | float | Qualcomm® SA8295P | 7.934 ms | 0 - 101 MB | NPU
90
- | CavaFace | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.589 ms | 0 - 102 MB | NPU
91
- | CavaFace | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.233 ms | 0 - 101 MB | NPU
92
 
93
  ## License
94
  * The license for the original implementation of CavaFace can be found
 
16
  A PyTorch-based framework for training face recognition models that generates facial embeddings for verification and identification tasks
17
 
18
  This is based on the implementation of CavaFace found [here](https://github.com/cavalleria/cavaface).
19
+ 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/tree/v0.49.1/qai_hub_models/models/cavaface) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
20
 
21
  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.
22
 
 
29
 
30
  | Runtime | Precision | Chipset | SDK Versions | Download |
31
  |---|---|---|---|---|
32
+ | 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/cavaface/releases/v0.49.1/cavaface-onnx-float.zip)
33
+ | QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/cavaface/releases/v0.49.1/cavaface-qnn_dlc-float.zip)
34
+ | 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/cavaface/releases/v0.49.1/cavaface-tflite-float.zip)
35
 
36
  For more device-specific assets and performance metrics, visit **[CavaFace on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/cavaface)**.
37
 
38
 
39
  ### Option 2: Export with Custom Configurations
40
 
41
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/cavaface) Python library to compile and export the model with your own:
42
  - Custom weights (e.g., fine-tuned checkpoints)
43
  - Custom input shapes
44
  - Target device and runtime configurations
45
 
46
  This option is ideal if you need to customize the model beyond the default configuration provided here.
47
 
48
+ See our repository for [CavaFace on GitHub](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/cavaface) for usage instructions.
49
 
50
  ## Model Details
51
 
 
60
  ## Performance Summary
61
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
62
  |---|---|---|---|---|---|---
63
+ | CavaFace | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.252 ms | 0 - 92 MB | NPU
64
+ | CavaFace | ONNX | float | Snapdragon® X2 Elite | 2.34 ms | 126 - 126 MB | NPU
65
+ | CavaFace | ONNX | float | Snapdragon® X Elite | 4.494 ms | 126 - 126 MB | NPU
66
+ | CavaFace | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 3.197 ms | 0 - 108 MB | NPU
67
+ | CavaFace | ONNX | float | Qualcomm® QCS8550 (Proxy) | 4.351 ms | 0 - 131 MB | NPU
68
+ | CavaFace | ONNX | float | Qualcomm® QCS9075 | 6.791 ms | 0 - 3 MB | NPU
69
+ | CavaFace | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.626 ms | 0 - 82 MB | NPU
70
+ | CavaFace | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.242 ms | 0 - 83 MB | NPU
71
+ | CavaFace | QNN_DLC | float | Snapdragon® X2 Elite | 2.566 ms | 0 - 0 MB | NPU
72
+ | CavaFace | QNN_DLC | float | Snapdragon® X Elite | 4.487 ms | 0 - 0 MB | NPU
73
+ | CavaFace | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 3.19 ms | 0 - 102 MB | NPU
74
+ | CavaFace | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 24.695 ms | 0 - 81 MB | NPU
75
+ | CavaFace | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 4.301 ms | 0 - 2 MB | NPU
76
+ | CavaFace | QNN_DLC | float | Qualcomm® SA8775P | 6.923 ms | 0 - 81 MB | NPU
77
+ | CavaFace | QNN_DLC | float | Qualcomm® QCS9075 | 6.769 ms | 0 - 2 MB | NPU
78
+ | CavaFace | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 8.861 ms | 0 - 109 MB | NPU
79
+ | CavaFace | QNN_DLC | float | Qualcomm® SA7255P | 24.695 ms | 0 - 81 MB | NPU
80
+ | CavaFace | QNN_DLC | float | Qualcomm® SA8295P | 7.964 ms | 0 - 85 MB | NPU
81
+ | CavaFace | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.611 ms | 0 - 83 MB | NPU
82
+ | CavaFace | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.217 ms | 0 - 101 MB | NPU
83
+ | CavaFace | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 3.145 ms | 0 - 216 MB | NPU
84
+ | CavaFace | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 24.553 ms | 0 - 98 MB | NPU
85
+ | CavaFace | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 4.227 ms | 0 - 2 MB | NPU
86
+ | CavaFace | TFLITE | float | Qualcomm® SA8775P | 6.895 ms | 0 - 98 MB | NPU
87
+ | CavaFace | TFLITE | float | Qualcomm® QCS9075 | 6.702 ms | 0 - 129 MB | NPU
88
+ | CavaFace | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 8.791 ms | 0 - 218 MB | NPU
89
+ | CavaFace | TFLITE | float | Qualcomm® SA7255P | 24.553 ms | 0 - 98 MB | NPU
90
+ | CavaFace | TFLITE | float | Qualcomm® SA8295P | 7.932 ms | 0 - 102 MB | NPU
91
+ | CavaFace | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.595 ms | 0 - 100 MB | NPU
92
 
93
  ## License
94
  * The license for the original implementation of CavaFace can be found