HRNetFace: Optimized for Qualcomm Devices

Detects attributes (liveness, eye closeness, mask presence, glasses presence, sunglasses presence) that apply to a given face.

This is based on the implementation of HRNetFace 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 2.42, ONNX Runtime 1.24.1	Download
ONNX	w8a8	Universal	QAIRT 2.42, ONNX Runtime 1.24.1	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
QNN_DLC	w8a8	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.17.0	Download
TFLITE	w8a8	Universal	QAIRT 2.43, TFLite 2.17.0	Download

For more device-specific assets and performance metrics, visit HRNetFace 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 HRNetFace on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.object_detection

Model Stats:

• Model checkpoint: HR18-COFW.pth
• Input resolution: 256x256
• Number of parameters: 9.68M
• Model size (float): 36.87MB
• Model size (w8a8): 17.7 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
HRNetFace	ONNX	float	Snapdragon® X Elite	3.326 ms	30 - 30 MB	NPU
HRNetFace	ONNX	float	Snapdragon® 8 Gen 3 Mobile	2.152 ms	0 - 148 MB	NPU
HRNetFace	ONNX	float	Qualcomm® QCS8550 (Proxy)	2.971 ms	0 - 34 MB	NPU
HRNetFace	ONNX	float	Qualcomm® QCS9075	4.743 ms	2 - 4 MB	NPU
HRNetFace	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1.652 ms	0 - 95 MB	NPU
HRNetFace	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.328 ms	1 - 98 MB	NPU
HRNetFace	ONNX	float	Snapdragon® X2 Elite	1.541 ms	31 - 31 MB	NPU
HRNetFace	ONNX	w8a8	Snapdragon® X Elite	1.552 ms	15 - 15 MB	NPU
HRNetFace	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.903 ms	0 - 153 MB	NPU
HRNetFace	ONNX	w8a8	Qualcomm® QCS6490	92.127 ms	18 - 36 MB	CPU
HRNetFace	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	1.304 ms	0 - 21 MB	NPU
HRNetFace	ONNX	w8a8	Qualcomm® QCS9075	1.548 ms	0 - 3 MB	NPU
HRNetFace	ONNX	w8a8	Qualcomm® QCM6690	49.641 ms	18 - 36 MB	CPU
HRNetFace	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.696 ms	0 - 95 MB	NPU
HRNetFace	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	47.06 ms	19 - 37 MB	CPU
HRNetFace	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.58 ms	0 - 105 MB	NPU
HRNetFace	ONNX	w8a8	Snapdragon® X2 Elite	0.652 ms	15 - 15 MB	NPU
HRNetFace	QNN_DLC	float	Snapdragon® X Elite	3.733 ms	1 - 1 MB	NPU
HRNetFace	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.373 ms	0 - 123 MB	NPU
HRNetFace	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	15.467 ms	1 - 76 MB	NPU
HRNetFace	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	3.351 ms	1 - 8 MB	NPU
HRNetFace	QNN_DLC	float	Qualcomm® SA8775P	5.21 ms	1 - 77 MB	NPU
HRNetFace	QNN_DLC	float	Qualcomm® QCS9075	5.106 ms	3 - 5 MB	NPU
HRNetFace	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	4.917 ms	0 - 105 MB	NPU
HRNetFace	QNN_DLC	float	Qualcomm® SA7255P	15.467 ms	1 - 76 MB	NPU
HRNetFace	QNN_DLC	float	Qualcomm® SA8295P	5.664 ms	1 - 61 MB	NPU
HRNetFace	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.789 ms	0 - 78 MB	NPU
HRNetFace	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.391 ms	1 - 81 MB	NPU
HRNetFace	QNN_DLC	float	Snapdragon® X2 Elite	1.907 ms	1 - 1 MB	NPU
HRNetFace	QNN_DLC	w8a8	Snapdragon® X Elite	1.573 ms	0 - 0 MB	NPU
HRNetFace	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.952 ms	0 - 112 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® QCS6490	3.898 ms	0 - 2 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	3.291 ms	0 - 70 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	1.367 ms	0 - 16 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® SA8775P	1.725 ms	0 - 72 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® QCS9075	1.646 ms	2 - 4 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® QCM6690	10.471 ms	0 - 198 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	1.814 ms	0 - 106 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® SA7255P	3.291 ms	0 - 70 MB	NPU
HRNetFace	QNN_DLC	w8a8	Qualcomm® SA8295P	2.241 ms	0 - 68 MB	NPU
HRNetFace	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.677 ms	0 - 72 MB	NPU
HRNetFace	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.639 ms	0 - 78 MB	NPU
HRNetFace	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.544 ms	0 - 72 MB	NPU
HRNetFace	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.896 ms	0 - 0 MB	NPU
HRNetFace	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.146 ms	0 - 143 MB	NPU
HRNetFace	TFLITE	float	Qualcomm® QCS8275 (Proxy)	15.232 ms	0 - 92 MB	NPU
HRNetFace	TFLITE	float	Qualcomm® QCS8550 (Proxy)	3.002 ms	0 - 3 MB	NPU
HRNetFace	TFLITE	float	Qualcomm® SA8775P	4.86 ms	0 - 93 MB	NPU
HRNetFace	TFLITE	float	Qualcomm® QCS9075	4.734 ms	0 - 35 MB	NPU
HRNetFace	TFLITE	float	Qualcomm® QCS8450 (Proxy)	4.496 ms	0 - 125 MB	NPU
HRNetFace	TFLITE	float	Qualcomm® SA7255P	15.232 ms	0 - 92 MB	NPU
HRNetFace	TFLITE	float	Qualcomm® SA8295P	5.273 ms	1 - 75 MB	NPU
HRNetFace	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1.667 ms	0 - 87 MB	NPU
HRNetFace	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.301 ms	0 - 96 MB	NPU
HRNetFace	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.836 ms	0 - 115 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® QCS6490	3.586 ms	0 - 18 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	3.194 ms	0 - 71 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	1.276 ms	0 - 3 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® SA8775P	1.683 ms	0 - 72 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® QCS9075	1.487 ms	0 - 18 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® QCM6690	9.845 ms	0 - 196 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	1.665 ms	0 - 110 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® SA7255P	3.194 ms	0 - 71 MB	NPU
HRNetFace	TFLITE	w8a8	Qualcomm® SA8295P	2.174 ms	0 - 68 MB	NPU
HRNetFace	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.69 ms	0 - 71 MB	NPU
HRNetFace	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.531 ms	0 - 75 MB	NPU
HRNetFace	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.574 ms	0 - 73 MB	NPU

License

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

References

• Deep High-Resolution Representation Learning for Visual Recognition
• Source Model Implementation

Community

• Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
• For questions or feedback please reach out to us.