SixDRepNet: Optimized for Qualcomm Devices

6DRepNet predicts head pose (pitch, yaw, roll) from a face image using a RepVGG-B1g2 backbone and a continuous 6D rotation representation, achieving robust and accurate head pose estimation.

This is based on the implementation of SixDRepNet 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 1.25.0	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

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

Model Details

Model Type: Model_use_case.pose_estimation

Model Stats:

• Input resolution: 224x224
• Number of parameters: 15.3M
• Model size (float): 58.4 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
face_detector	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.49 ms	0 - 166 MB	NPU
face_detector	ONNX	float	Snapdragon® X2 Elite	1.607 ms	208 - 208 MB	NPU
face_detector	ONNX	float	Snapdragon® X Elite	3.644 ms	176 - 176 MB	NPU
face_detector	ONNX	float	Snapdragon® 8 Gen 3 Mobile	2.24 ms	0 - 174 MB	NPU
face_detector	ONNX	float	Qualcomm® QCS8550 (Proxy)	3.541 ms	0 - 42 MB	NPU
face_detector	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1.809 ms	0 - 164 MB	NPU
face_detector	ONNX	float	Qualcomm® QCS9075	5.335 ms	5 - 50 MB	NPU
face_detector	ONNX	float	Qualcomm® QCS8750	1.809 ms	0 - 164 MB	NPU
face_detector	ONNX	float	Qualcomm® QCS7181	3.644 ms	176 - 176 MB	NPU
face_detector	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	5.425 ms	5 - 158 MB	NPU
face_detector	QNN_DLC	float	Snapdragon® X2 Elite	5.832 ms	5 - 5 MB	NPU
face_detector	QNN_DLC	float	Snapdragon® X Elite	16.701 ms	5 - 5 MB	NPU
face_detector	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	9.303 ms	5 - 172 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® QCS8275	28.301 ms	1 - 151 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	15.574 ms	5 - 88 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® SA8775P	16.569 ms	1 - 153 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® SA8650P	16.569 ms	1 - 153 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® SA8255P	16.569 ms	1 - 153 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	24.66 ms	5 - 175 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® SA7255P	28.301 ms	1 - 151 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® SA8295P	20.599 ms	0 - 150 MB	NPU
face_detector	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	6.916 ms	0 - 150 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® QCS9075	19.394 ms	5 - 12 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® QCS8750	6.916 ms	0 - 150 MB	NPU
face_detector	QNN_DLC	float	Qualcomm® QCS7181	16.701 ms	5 - 5 MB	NPU
face_detector	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	5.472 ms	1 - 155 MB	NPU
face_detector	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	9.279 ms	0 - 166 MB	NPU
face_detector	TFLITE	float	Qualcomm® QCS8275	28.329 ms	1 - 150 MB	NPU
face_detector	TFLITE	float	Qualcomm® QCS8550 (Proxy)	15.649 ms	1 - 3 MB	NPU
face_detector	TFLITE	float	Qualcomm® SA8775P	16.518 ms	1 - 154 MB	NPU
face_detector	TFLITE	float	Qualcomm® SA8650P	16.518 ms	1 - 154 MB	NPU
face_detector	TFLITE	float	Qualcomm® SA8255P	16.518 ms	1 - 154 MB	NPU
face_detector	TFLITE	float	Qualcomm® QCS8450 (Proxy)	24.642 ms	1 - 173 MB	NPU
face_detector	TFLITE	float	Qualcomm® SA7255P	28.329 ms	1 - 150 MB	NPU
face_detector	TFLITE	float	Qualcomm® SA8295P	20.681 ms	1 - 151 MB	NPU
face_detector	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	6.928 ms	0 - 149 MB	NPU
face_detector	TFLITE	float	Qualcomm® QCS9075	19.232 ms	0 - 9 MB	NPU
face_detector	TFLITE	float	Qualcomm® QCS8750	6.928 ms	0 - 149 MB	NPU
pose_estimator	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.329 ms	1 - 27 MB	NPU
pose_estimator	ONNX	float	Snapdragon® X2 Elite	1.295 ms	211 - 211 MB	NPU
pose_estimator	ONNX	float	Snapdragon® X Elite	2.491 ms	156 - 156 MB	NPU
pose_estimator	ONNX	float	Snapdragon® 8 Gen 3 Mobile	2.062 ms	0 - 41 MB	NPU
pose_estimator	ONNX	float	Qualcomm® QCS8550 (Proxy)	2.632 ms	0 - 374 MB	NPU
pose_estimator	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1.637 ms	0 - 24 MB	NPU
pose_estimator	ONNX	float	Qualcomm® QCS9075	4.56 ms	1 - 46 MB	NPU
pose_estimator	ONNX	float	Qualcomm® QCS8750	1.637 ms	0 - 24 MB	NPU
pose_estimator	ONNX	float	Qualcomm® QCS7181	2.491 ms	156 - 156 MB	NPU
pose_estimator	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.42 ms	1 - 29 MB	NPU
pose_estimator	QNN_DLC	float	Snapdragon® X2 Elite	1.517 ms	1 - 1 MB	NPU
pose_estimator	QNN_DLC	float	Snapdragon® X Elite	2.855 ms	1 - 1 MB	NPU
pose_estimator	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.241 ms	0 - 39 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® QCS8275	17.824 ms	1 - 25 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	2.774 ms	1 - 2 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® SA8775P	4.793 ms	0 - 26 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® SA8650P	4.793 ms	0 - 26 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® SA8255P	4.793 ms	0 - 26 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	6.534 ms	0 - 39 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® SA7255P	17.824 ms	1 - 25 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® SA8295P	5.402 ms	1 - 22 MB	NPU
pose_estimator	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.684 ms	1 - 27 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® QCS9075	4.905 ms	3 - 5 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® QCS8750	1.684 ms	1 - 27 MB	NPU
pose_estimator	QNN_DLC	float	Qualcomm® QCS7181	2.855 ms	1 - 1 MB	NPU
pose_estimator	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.366 ms	0 - 30 MB	NPU
pose_estimator	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.249 ms	0 - 44 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® QCS8275	17.402 ms	0 - 26 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® QCS8550 (Proxy)	2.848 ms	0 - 2 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® SA8775P	4.799 ms	0 - 27 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® SA8650P	4.799 ms	0 - 27 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® SA8255P	4.799 ms	0 - 27 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® QCS8450 (Proxy)	6.509 ms	0 - 47 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® SA7255P	17.402 ms	0 - 26 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® SA8295P	5.346 ms	0 - 27 MB	NPU
pose_estimator	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1.707 ms	0 - 30 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® QCS9075	4.743 ms	0 - 78 MB	NPU
pose_estimator	TFLITE	float	Qualcomm® QCS8750	1.707 ms	0 - 30 MB	NPU

License

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

References

• 6D Rotation Representation for Unconstrained Head Pose Estimation
• 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.