Hugging Face's logo

qualcomm
/
HRNetPose

Keypoint Detection
PyTorch
bu_auto
android
Model card Files
xet
 Community

HRNetPose: Optimized for Qualcomm Devices

HRNet performs pose estimation in high-resolution representations.

This is based on the implementation of HRNetPose found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up to run these models on a hosted Qualcomm® device.

Getting Started

There are two ways to deploy this model on your device:

Option 1: Download Pre-Exported Models

Below are pre-exported model assets ready for deployment.

Runtime Precision Chipset SDK Versions Download
ONNX float Universal QAIRT 2.50, ONNX Runtime 1.27.1 Download
ONNX w8a16 Universal QAIRT 2.50, ONNX Runtime 1.27.1 Download
ONNX w8a8 Universal QAIRT 2.50, ONNX Runtime 1.27.1 Download
QNN_DLC float Universal QAIRT 2.50 Download
QNN_DLC w8a16 Universal QAIRT 2.50 Download
QNN_DLC w8a8 Universal QAIRT 2.50 Download
TFLITE float Universal QAIRT 2.50 Download
TFLITE w8a8 Universal QAIRT 2.50 Download

For more device-specific assets and performance metrics, visit HRNetPose on Qualcomm® AI Hub.

Option 2: Export with Custom Configurations

Use the Qualcomm® AI Hub Models Python library to compile and export the model with your own:

  • Custom weights (e.g., fine-tuned checkpoints)
  • Custom input shapes
  • Target device and runtime configurations

This option is ideal if you need to customize the model beyond the default configuration provided here.

See our repository for HRNetPose on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.pose_estimation

Model Stats:

  • Input resolution: 256x192
  • Model checkpoint: hrnet_posenet_FP32_state_dict
  • Model size (float): 109 MB
  • Model size (w8a8): 28.1 MB
  • Number of parameters: 28.5M

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
HRNetPose ONNX float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 1.265 ms 1 - 93 MB NPU
HRNetPose ONNX float Snapdragon® 8 Elite For Galaxy Mobile 1.456 ms 0 - 91 MB NPU
HRNetPose ONNX float Snapdragon® X2 Elite 1.282 ms 2 - 2 MB NPU
HRNetPose ONNX float Snapdragon® X Elite 2.373 ms 56 - 56 MB NPU
HRNetPose ONNX float Snapdragon® 8 Gen 3 Mobile 1.807 ms 0 - 144 MB NPU
HRNetPose ONNX float Snapdragon® 8 Gen 1 Mobile 3.734 ms 1 - 134 MB NPU
HRNetPose ONNX float Qualcomm® Dragonwing™ IQ-8275 4.324 ms 0 - 5 MB NPU
HRNetPose ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 2.375 ms 0 - 58 MB NPU
HRNetPose ONNX float Qualcomm® QCS8450 3.734 ms 1 - 134 MB NPU
HRNetPose ONNX float Qualcomm® Dragonwing™ IQ-9075 3.871 ms 0 - 4 MB NPU
HRNetPose ONNX float Qualcomm® Dragonwing™ IQ-X7181 2.373 ms 56 - 56 MB NPU
HRNetPose ONNX float Qualcomm® Dragonwing™ Q-8750 1.456 ms 0 - 91 MB NPU
HRNetPose ONNX w8a16 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.819 ms 0 - 140 MB NPU
HRNetPose ONNX w8a16 Snapdragon® 8 Elite For Galaxy Mobile 0.956 ms 0 - 137 MB NPU
HRNetPose ONNX w8a16 Snapdragon® X2 Elite 1.008 ms 1 - 1 MB NPU
HRNetPose ONNX w8a16 Snapdragon® X Elite 1.714 ms 28 - 28 MB NPU
HRNetPose ONNX w8a16 Snapdragon® 8 Gen 3 Mobile 1.229 ms 0 - 203 MB NPU
HRNetPose ONNX w8a16 Snapdragon® 8 Gen 1 Mobile 2.175 ms 0 - 206 MB NPU
HRNetPose ONNX w8a16 Qualcomm® Dragonwing™ QCS6490 6.459 ms 0 - 4 MB NPU
HRNetPose ONNX w8a16 Qualcomm® Dragonwing™ IQ-8275 1.942 ms 0 - 4 MB NPU
HRNetPose ONNX w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 1.713 ms 0 - 35 MB NPU
HRNetPose ONNX w8a16 Qualcomm® QCS8450 2.175 ms 0 - 206 MB NPU
HRNetPose ONNX w8a16 Qualcomm® Dragonwing™ IQ-9075 1.896 ms 0 - 4 MB NPU
HRNetPose ONNX w8a16 Qualcomm® Dragonwing™ IQ-X7181 1.714 ms 28 - 28 MB NPU
HRNetPose ONNX w8a16 Qualcomm® Dragonwing™ Q-6690 22.571 ms 0 - 243 MB NPU
HRNetPose ONNX w8a16 Qualcomm® Dragonwing™ Q-7790 2.34 ms 0 - 143 MB NPU
HRNetPose ONNX w8a16 Qualcomm® Dragonwing™ Q-8750 0.956 ms 0 - 137 MB NPU
HRNetPose ONNX w8a16 Snapdragon® 7 Gen 4 Mobile 2.34 ms 0 - 143 MB NPU
HRNetPose ONNX w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.645 ms 0 - 130 MB NPU
HRNetPose ONNX w8a8 Snapdragon® 8 Elite For Galaxy Mobile 0.703 ms 0 - 126 MB NPU
HRNetPose ONNX w8a8 Snapdragon® X2 Elite 0.539 ms 1 - 1 MB NPU
HRNetPose ONNX w8a8 Snapdragon® X Elite 1.066 ms 28 - 28 MB NPU
HRNetPose ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 0.848 ms 0 - 190 MB NPU
HRNetPose ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 1.408 ms 0 - 192 MB NPU
HRNetPose ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 3.438 ms 0 - 4 MB NPU
HRNetPose ONNX w8a8 Qualcomm® Dragonwing™ IQ-8275 1.358 ms 0 - 4 MB NPU
HRNetPose ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 1.14 ms 0 - 129 MB NPU
HRNetPose ONNX w8a8 Qualcomm® QCS8450 1.408 ms 0 - 192 MB NPU
HRNetPose ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 1.364 ms 0 - 3 MB NPU
HRNetPose ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 1.066 ms 28 - 28 MB NPU
HRNetPose ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 9.998 ms 0 - 237 MB NPU
HRNetPose ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 1.434 ms 0 - 127 MB NPU
HRNetPose ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 0.703 ms 0 - 126 MB NPU
HRNetPose ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 1.434 ms 0 - 127 MB NPU
HRNetPose QNN_DLC float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 1.34 ms 1 - 78 MB NPU
HRNetPose QNN_DLC float Snapdragon® 8 Elite For Galaxy Mobile 1.532 ms 0 - 75 MB NPU
HRNetPose QNN_DLC float Snapdragon® X2 Elite 1.751 ms 1 - 1 MB NPU
HRNetPose QNN_DLC float Snapdragon® X Elite 2.871 ms 1 - 1 MB NPU
HRNetPose QNN_DLC float Snapdragon® 8 Gen 3 Mobile 2.002 ms 0 - 117 MB NPU
HRNetPose QNN_DLC float Snapdragon® 8 Gen 1 Mobile 4.386 ms 0 - 104 MB NPU
HRNetPose QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 4.347 ms 1 - 4 MB NPU
HRNetPose QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 2.677 ms 1 - 125 MB NPU
HRNetPose QNN_DLC float Qualcomm® QCS8450 4.386 ms 0 - 104 MB NPU
HRNetPose QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 4.024 ms 3 - 6 MB NPU
HRNetPose QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 2.871 ms 1 - 1 MB NPU
HRNetPose QNN_DLC float Qualcomm® Dragonwing™ Q-8750 1.532 ms 0 - 75 MB NPU
HRNetPose QNN_DLC float Qualcomm® SA8295P 4.453 ms 1 - 60 MB NPU
HRNetPose QNN_DLC w8a16 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.844 ms 0 - 107 MB NPU
HRNetPose QNN_DLC w8a16 Snapdragon® 8 Elite For Galaxy Mobile 1.033 ms 0 - 108 MB NPU
HRNetPose QNN_DLC w8a16 Snapdragon® X2 Elite 1.3 ms 0 - 0 MB NPU
HRNetPose QNN_DLC w8a16 Snapdragon® X Elite 2.115 ms 0 - 0 MB NPU
HRNetPose QNN_DLC w8a16 Snapdragon® 8 Gen 3 Mobile 1.385 ms 0 - 151 MB NPU
HRNetPose QNN_DLC w8a16 Snapdragon® 8 Gen 1 Mobile 2.362 ms 0 - 152 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® Dragonwing™ QCS6490 7.009 ms 0 - 3 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-8275 2.083 ms 0 - 4 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 1.904 ms 0 - 2 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® SA8775P 2.207 ms 0 - 107 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® SA8650P 2.207 ms 0 - 107 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® SA8255P 2.207 ms 0 - 107 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® QCS8450 2.362 ms 0 - 152 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-9075 2.059 ms 0 - 3 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-X7181 2.115 ms 0 - 0 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-6690 19.8 ms 0 - 211 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-7790 2.552 ms 0 - 104 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-8750 1.033 ms 0 - 108 MB NPU
HRNetPose QNN_DLC w8a16 Qualcomm® SA8295P 2.924 ms 0 - 102 MB NPU
HRNetPose QNN_DLC w8a16 Snapdragon® 7 Gen 4 Mobile 2.552 ms 0 - 104 MB NPU
HRNetPose QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.573 ms 0 - 96 MB NPU
HRNetPose QNN_DLC w8a8 Snapdragon® 8 Elite For Galaxy Mobile 0.621 ms 0 - 93 MB NPU
HRNetPose QNN_DLC w8a8 Snapdragon® X2 Elite 0.693 ms 0 - 0 MB NPU
HRNetPose QNN_DLC w8a8 Snapdragon® X Elite 1.227 ms 0 - 0 MB NPU
HRNetPose QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 0.823 ms 0 - 138 MB NPU
HRNetPose QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 1.4 ms 0 - 140 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS6490 3.736 ms 2 - 4 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 1.284 ms 0 - 3 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 1.115 ms 0 - 2 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® SA8650P 1.494 ms 0 - 92 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® SA8255P 1.494 ms 0 - 92 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® QCS8450 1.4 ms 0 - 140 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 1.318 ms 0 - 2 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 1.227 ms 0 - 0 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-6690 10.64 ms 0 - 207 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-7790 1.491 ms 0 - 94 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 0.621 ms 0 - 93 MB NPU
HRNetPose QNN_DLC w8a8 Qualcomm® SA8295P 1.82 ms 0 - 93 MB NPU
HRNetPose QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 1.491 ms 0 - 94 MB NPU
HRNetPose TFLITE float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 1.351 ms 0 - 119 MB NPU
HRNetPose TFLITE float Snapdragon® 8 Elite For Galaxy Mobile 1.567 ms 0 - 108 MB NPU
HRNetPose TFLITE float Snapdragon® 8 Gen 3 Mobile 1.979 ms 0 - 187 MB NPU
HRNetPose TFLITE float Snapdragon® 8 Gen 1 Mobile 4.396 ms 0 - 175 MB NPU
HRNetPose TFLITE float Qualcomm® Dragonwing™ IQ-8275 4.38 ms 0 - 59 MB NPU
HRNetPose TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 2.67 ms 0 - 91 MB NPU
HRNetPose TFLITE float Qualcomm® SA8775P 4.242 ms 0 - 115 MB NPU
HRNetPose TFLITE float Qualcomm® SA8650P 4.242 ms 0 - 115 MB NPU
HRNetPose TFLITE float Qualcomm® SA8255P 4.242 ms 0 - 115 MB NPU
HRNetPose TFLITE float Qualcomm® QCS8450 4.396 ms 0 - 175 MB NPU
HRNetPose TFLITE float Qualcomm® Dragonwing™ IQ-9075 4.03 ms 0 - 58 MB NPU
HRNetPose TFLITE float Qualcomm® Dragonwing™ Q-8750 1.567 ms 0 - 108 MB NPU
HRNetPose TFLITE float Qualcomm® SA8295P 4.452 ms 9 - 111 MB NPU
HRNetPose TFLITE w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.516 ms 0 - 97 MB NPU
HRNetPose TFLITE w8a8 Snapdragon® 8 Elite For Galaxy Mobile 0.587 ms 0 - 96 MB NPU
HRNetPose TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 0.694 ms 0 - 140 MB NPU
HRNetPose TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 1.16 ms 0 - 149 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 3.518 ms 0 - 30 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 1.133 ms 0 - 31 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 0.936 ms 0 - 2 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® SA8650P 1.357 ms 0 - 90 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® SA8255P 1.357 ms 0 - 90 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® QCS8450 1.16 ms 0 - 149 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 1.157 ms 0 - 30 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 9.657 ms 0 - 202 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 1.339 ms 0 - 90 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 0.587 ms 0 - 96 MB NPU
HRNetPose TFLITE w8a8 Qualcomm® SA7255P 91.19 ms 1 - 43 MB GPU
HRNetPose TFLITE w8a8 Qualcomm® SA8295P 1.642 ms 0 - 89 MB NPU
HRNetPose TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 1.339 ms 0 - 90 MB NPU

License

  • The license for the original implementation of HRNetPose can be found here.

References

Community

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Keypoint Detection
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Paper for qualcomm/HRNetPose