Hugging Face's logo

qualcomm
/
MediaPipe-Hand-Detection

Object Detection
PyTorch
real_time
android
Model card Files
xet
 Community
1
qaihm-bot commited on
Commit
00ea5bb
·
verified ·
1 Parent(s): de2f56f

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 +62 -62
README.md CHANGED
@@ -15,7 +15,7 @@ pipeline_tag: object-detection
15
  The MediaPipe Hand Landmark Detector is a machine learning pipeline that predicts bounding boxes and pose skeletons of hands in an image.
16
 
17
  This is based on the implementation of MediaPipe-Hand-Detection found [here](https://github.com/zmurez/MediaPipePyTorch/).
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/mediapipe_hand) 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/mediapipe_hand/releases/v0.48.0/mediapipe_hand-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/mediapipe_hand/releases/v0.48.0/mediapipe_hand-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/mediapipe_hand/releases/v0.48.0/mediapipe_hand-tflite-float.zip)
34
 
35
  For more device-specific assets and performance metrics, visit **[MediaPipe-Hand-Detection on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/mediapipe_hand)**.
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/mediapipe_hand) 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 [MediaPipe-Hand-Detection on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/mediapipe_hand) for usage instructions.
48
 
49
  ## Model Details
50
 
@@ -60,64 +60,64 @@ See our repository for [MediaPipe-Hand-Detection on GitHub](https://github.com/q
60
  ## Performance Summary
61
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
62
  |---|---|---|---|---|---|---
63
- | HandDetector | ONNX | float | Snapdragon® X2 Elite | 0.487 ms | 0 - 0 MB | NPU
64
- | HandDetector | ONNX | float | Snapdragon® X Elite | 1.014 ms | 3 - 3 MB | NPU
65
- | HandDetector | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.598 ms | 0 - 65 MB | NPU
66
- | HandDetector | ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.839 ms | 0 - 14 MB | NPU
67
- | HandDetector | ONNX | float | Qualcomm® QCS9075 | 1.275 ms | 1 - 3 MB | NPU
68
- | HandDetector | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.495 ms | 0 - 38 MB | NPU
69
- | HandDetector | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.419 ms | 0 - 47 MB | NPU
70
- | HandDetector | QNN_DLC | float | Snapdragon® X2 Elite | 0.617 ms | 1 - 1 MB | NPU
71
- | HandDetector | QNN_DLC | float | Snapdragon® X Elite | 0.908 ms | 1 - 1 MB | NPU
72
- | HandDetector | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.534 ms | 0 - 59 MB | NPU
73
- | HandDetector | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 3.782 ms | 1 - 42 MB | NPU
74
- | HandDetector | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 0.733 ms | 0 - 10 MB | NPU
75
- | HandDetector | QNN_DLC | float | Qualcomm® SA8775P | 1.297 ms | 1 - 43 MB | NPU
 
76
  | HandDetector | QNN_DLC | float | Qualcomm® QCS9075 | 1.13 ms | 1 - 3 MB | NPU
77
- | HandDetector | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 1.349 ms | 0 - 53 MB | NPU
78
- | HandDetector | QNN_DLC | float | Qualcomm® SA7255P | 3.782 ms | 1 - 42 MB | NPU
79
- | HandDetector | QNN_DLC | float | Qualcomm® SA8295P | 1.673 ms | 0 - 31 MB | NPU
80
- | HandDetector | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.445 ms | 0 - 43 MB | NPU
81
- | HandDetector | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.381 ms | 1 - 44 MB | NPU
82
- | HandDetector | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.536 ms | 0 - 58 MB | NPU
83
- | HandDetector | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 3.836 ms | 0 - 38 MB | NPU
84
- | HandDetector | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 0.74 ms | 0 - 4 MB | NPU
85
- | HandDetector | TFLITE | float | Qualcomm® SA8775P | 1.338 ms | 0 - 42 MB | NPU
86
- | HandDetector | TFLITE | float | Qualcomm® QCS9075 | 1.138 ms | 0 - 7 MB | NPU
87
- | HandDetector | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 1.368 ms | 0 - 50 MB | NPU
88
- | HandDetector | TFLITE | float | Qualcomm® SA7255P | 3.836 ms | 0 - 38 MB | NPU
89
- | HandDetector | TFLITE | float | Qualcomm® SA8295P | 1.714 ms | 0 - 31 MB | NPU
90
- | HandDetector | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.452 ms | 0 - 36 MB | NPU
91
- | HandDetector | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.379 ms | 0 - 42 MB | NPU
92
- | HandLandmarkDetector | ONNX | float | Snapdragon® X2 Elite | 0.719 ms | 6 - 6 MB | NPU
93
- | HandLandmarkDetector | ONNX | float | Snapdragon® X Elite | 1.386 ms | 6 - 6 MB | NPU
94
- | HandLandmarkDetector | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.832 ms | 0 - 69 MB | NPU
95
- | HandLandmarkDetector | ONNX | float | Qualcomm® QCS8550 (Proxy) | 1.1 ms | 0 - 11 MB | NPU
96
- | HandLandmarkDetector | ONNX | float | Qualcomm® QCS9075 | 1.83 ms | 1 - 4 MB | NPU
97
- | HandLandmarkDetector | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.661 ms | 0 - 45 MB | NPU
98
- | HandLandmarkDetector | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.559 ms | 0 - 43 MB | NPU
99
- | HandLandmarkDetector | QNN_DLC | float | Snapdragon® X2 Elite | 0.876 ms | 1 - 1 MB | NPU
100
- | HandLandmarkDetector | QNN_DLC | float | Snapdragon® X Elite | 1.283 ms | 1 - 1 MB | NPU
101
- | HandLandmarkDetector | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.754 ms | 0 - 57 MB | NPU
102
- | HandLandmarkDetector | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 5.336 ms | 1 - 35 MB | NPU
103
- | HandLandmarkDetector | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 1.005 ms | 0 - 7 MB | NPU
104
- | HandLandmarkDetector | QNN_DLC | float | Qualcomm® SA8775P | 1.893 ms | 1 - 39 MB | NPU
105
- | HandLandmarkDetector | QNN_DLC | float | Qualcomm® QCS9075 | 1.692 ms | 3 - 5 MB | NPU
106
- | HandLandmarkDetector | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 1.947 ms | 0 - 52 MB | NPU
107
- | HandLandmarkDetector | QNN_DLC | float | Qualcomm® SA7255P | 5.336 ms | 1 - 35 MB | NPU
108
- | HandLandmarkDetector | QNN_DLC | float | Qualcomm® SA8295P | 2.22 ms | 0 - 30 MB | NPU
109
  | HandLandmarkDetector | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.601 ms | 0 - 34 MB | NPU
110
- | HandLandmarkDetector | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.513 ms | 1 - 39 MB | NPU
111
- | HandLandmarkDetector | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.764 ms | 0 - 60 MB | NPU
112
- | HandLandmarkDetector | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 5.351 ms | 0 - 41 MB | NPU
113
- | HandLandmarkDetector | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 1.028 ms | 0 - 2 MB | NPU
114
- | HandLandmarkDetector | TFLITE | float | Qualcomm® SA8775P | 1.89 ms | 0 - 44 MB | NPU
115
- | HandLandmarkDetector | TFLITE | float | Qualcomm® QCS9075 | 1.679 ms | 0 - 9 MB | NPU
116
- | HandLandmarkDetector | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 1.97 ms | 0 - 57 MB | NPU
117
- | HandLandmarkDetector | TFLITE | float | Qualcomm® SA7255P | 5.351 ms | 0 - 41 MB | NPU
118
- | HandLandmarkDetector | TFLITE | float | Qualcomm® SA8295P | 2.232 ms | 0 - 35 MB | NPU
119
- | HandLandmarkDetector | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.605 ms | 0 - 43 MB | NPU
120
- | HandLandmarkDetector | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.51 ms | 0 - 44 MB | NPU
121
 
122
  ## License
123
  * The license for the original implementation of MediaPipe-Hand-Detection can be found
 
15
  The MediaPipe Hand Landmark Detector is a machine learning pipeline that predicts bounding boxes and pose skeletons of hands in an image.
16
 
17
  This is based on the implementation of MediaPipe-Hand-Detection found [here](https://github.com/zmurez/MediaPipePyTorch/).
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/tree/v0.49.1/qai_hub_models/models/mediapipe_hand) 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/mediapipe_hand/releases/v0.49.1/mediapipe_hand-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/mediapipe_hand/releases/v0.49.1/mediapipe_hand-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/mediapipe_hand/releases/v0.49.1/mediapipe_hand-tflite-float.zip)
34
 
35
  For more device-specific assets and performance metrics, visit **[MediaPipe-Hand-Detection on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/mediapipe_hand)**.
36
 
37
 
38
  ### Option 2: Export with Custom Configurations
39
 
40
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/mediapipe_hand) 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 [MediaPipe-Hand-Detection on GitHub](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/mediapipe_hand) for usage instructions.
48
 
49
  ## Model Details
50
 
 
60
  ## Performance Summary
61
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
62
  |---|---|---|---|---|---|---
63
+ | HandDetector | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.42 ms | 0 - 47 MB | NPU
64
+ | HandDetector | ONNX | float | Snapdragon® X2 Elite | 0.489 ms | 0 - 0 MB | NPU
65
+ | HandDetector | ONNX | float | Snapdragon® X Elite | 1.019 ms | 3 - 3 MB | NPU
66
+ | HandDetector | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.595 ms | 0 - 74 MB | NPU
67
+ | HandDetector | ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.843 ms | 0 - 7 MB | NPU
68
+ | HandDetector | ONNX | float | Qualcomm® QCS9075 | 1.267 ms | 1 - 3 MB | NPU
69
+ | HandDetector | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.499 ms | 0 - 49 MB | NPU
70
+ | HandDetector | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.378 ms | 1 - 44 MB | NPU
71
+ | HandDetector | QNN_DLC | float | Snapdragon® X2 Elite | 0.628 ms | 1 - 1 MB | NPU
72
+ | HandDetector | QNN_DLC | float | Snapdragon® X Elite | 0.926 ms | 1 - 1 MB | NPU
73
+ | HandDetector | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.537 ms | 0 - 60 MB | NPU
74
+ | HandDetector | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 3.819 ms | 1 - 40 MB | NPU
75
+ | HandDetector | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 0.736 ms | 1 - 2 MB | NPU
76
+ | HandDetector | QNN_DLC | float | Qualcomm® SA8775P | 5.169 ms | 1 - 40 MB | NPU
77
  | HandDetector | QNN_DLC | float | Qualcomm® QCS9075 | 1.13 ms | 1 - 3 MB | NPU
78
+ | HandDetector | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 1.389 ms | 0 - 54 MB | NPU
79
+ | HandDetector | QNN_DLC | float | Qualcomm® SA7255P | 3.819 ms | 1 - 40 MB | NPU
80
+ | HandDetector | QNN_DLC | float | Qualcomm® SA8295P | 1.691 ms | 0 - 31 MB | NPU
81
+ | HandDetector | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.446 ms | 0 - 40 MB | NPU
82
+ | HandDetector | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.381 ms | 0 - 42 MB | NPU
83
+ | HandDetector | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.535 ms | 0 - 58 MB | NPU
84
+ | HandDetector | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 3.831 ms | 0 - 38 MB | NPU
85
+ | HandDetector | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 0.738 ms | 0 - 2 MB | NPU
86
+ | HandDetector | TFLITE | float | Qualcomm® SA8775P | 1.323 ms | 0 - 42 MB | NPU
87
+ | HandDetector | TFLITE | float | Qualcomm® QCS9075 | 1.144 ms | 0 - 7 MB | NPU
88
+ | HandDetector | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 1.399 ms | 0 - 51 MB | NPU
89
+ | HandDetector | TFLITE | float | Qualcomm® SA7255P | 3.831 ms | 0 - 38 MB | NPU
90
+ | HandDetector | TFLITE | float | Qualcomm® SA8295P | 1.699 ms | 0 - 30 MB | NPU
91
+ | HandDetector | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.454 ms | 0 - 44 MB | NPU
92
+ | HandLandmarkDetector | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.563 ms | 0 - 43 MB | NPU
93
+ | HandLandmarkDetector | ONNX | float | Snapdragon® X2 Elite | 0.716 ms | 6 - 6 MB | NPU
94
+ | HandLandmarkDetector | ONNX | float | Snapdragon® X Elite | 1.413 ms | 6 - 6 MB | NPU
95
+ | HandLandmarkDetector | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.846 ms | 0 - 69 MB | NPU
96
+ | HandLandmarkDetector | ONNX | float | Qualcomm® QCS8550 (Proxy) | 1.113 ms | 0 - 9 MB | NPU
97
+ | HandLandmarkDetector | ONNX | float | Qualcomm® QCS9075 | 1.843 ms | 1 - 4 MB | NPU
98
+ | HandLandmarkDetector | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.659 ms | 0 - 44 MB | NPU
99
+ | HandLandmarkDetector | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.516 ms | 1 - 39 MB | NPU
100
+ | HandLandmarkDetector | QNN_DLC | float | Snapdragon® X2 Elite | 0.856 ms | 1 - 1 MB | NPU
101
+ | HandLandmarkDetector | QNN_DLC | float | Snapdragon® X Elite | 1.284 ms | 1 - 1 MB | NPU
102
+ | HandLandmarkDetector | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.769 ms | 0 - 61 MB | NPU
103
+ | HandLandmarkDetector | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 5.344 ms | 1 - 35 MB | NPU
104
+ | HandLandmarkDetector | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 1.007 ms | 1 - 2 MB | NPU
105
+ | HandLandmarkDetector | QNN_DLC | float | Qualcomm® SA8775P | 1.871 ms | 1 - 39 MB | NPU
106
+ | HandLandmarkDetector | QNN_DLC | float | Qualcomm® QCS9075 | 1.696 ms | 1 - 3 MB | NPU
107
+ | HandLandmarkDetector | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 1.934 ms | 0 - 52 MB | NPU
108
+ | HandLandmarkDetector | QNN_DLC | float | Qualcomm® SA7255P | 5.344 ms | 1 - 35 MB | NPU
109
+ | HandLandmarkDetector | QNN_DLC | float | Qualcomm® SA8295P | 2.245 ms | 0 - 31 MB | NPU
110
  | HandLandmarkDetector | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.601 ms | 0 - 34 MB | NPU
111
+ | HandLandmarkDetector | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.513 ms | 0 - 44 MB | NPU
112
+ | HandLandmarkDetector | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.776 ms | 0 - 60 MB | NPU
113
+ | HandLandmarkDetector | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 5.345 ms | 0 - 40 MB | NPU
114
+ | HandLandmarkDetector | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 1.0 ms | 0 - 79 MB | NPU
115
+ | HandLandmarkDetector | TFLITE | float | Qualcomm® SA8775P | 1.869 ms | 0 - 43 MB | NPU
116
+ | HandLandmarkDetector | TFLITE | float | Qualcomm® QCS9075 | 1.686 ms | 0 - 9 MB | NPU
117
+ | HandLandmarkDetector | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 1.948 ms | 0 - 57 MB | NPU
118
+ | HandLandmarkDetector | TFLITE | float | Qualcomm® SA7255P | 5.345 ms | 0 - 40 MB | NPU
119
+ | HandLandmarkDetector | TFLITE | float | Qualcomm® SA8295P | 2.242 ms | 0 - 35 MB | NPU
120
+ | HandLandmarkDetector | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.609 ms | 0 - 45 MB | NPU
 
121
 
122
  ## License
123
  * The license for the original implementation of MediaPipe-Hand-Detection can be found