Movenet: Optimized for Qualcomm Devices

Movenet performs pose estimation on human images.

This is based on the implementation of Movenet 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_mixed_int16	Universal	QAIRT 2.50, ONNX Runtime 1.27.1	Download
QNN_DLC	float	Universal	QAIRT 2.50	Download
QNN_DLC	w8a16_mixed_int16	Universal	QAIRT 2.50	Download
TFLITE	float	Universal	QAIRT 2.50	Download

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

Model Details

Model Type: Model_use_case.pose_estimation

Model Stats:

• Input resolution: 192x192
• Model checkpoint: None
• Model size (float): 8.91 MB
• Number of parameters: 2.33M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Movenet	ONNX	float	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.435 ms	0 - 37 MB	NPU
Movenet	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	0.553 ms	0 - 36 MB	NPU
Movenet	ONNX	float	Snapdragon® X2 Elite	0.469 ms	1 - 1 MB	NPU
Movenet	ONNX	float	Snapdragon® X Elite	1.04 ms	6 - 6 MB	NPU
Movenet	ONNX	float	Snapdragon® 8 Gen 3 Mobile	0.746 ms	0 - 55 MB	NPU
Movenet	ONNX	float	Snapdragon® 8 Gen 1 Mobile	1.22 ms	0 - 60 MB	NPU
Movenet	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	1.289 ms	0 - 4 MB	NPU
Movenet	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.987 ms	0 - 7 MB	NPU
Movenet	ONNX	float	Qualcomm® QCS8450	1.22 ms	0 - 60 MB	NPU
Movenet	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	1.286 ms	0 - 4 MB	NPU
Movenet	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	1.04 ms	6 - 6 MB	NPU
Movenet	ONNX	float	Qualcomm® Dragonwing™ Q-8750	0.553 ms	0 - 36 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.411 ms	0 - 47 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Snapdragon® 8 Elite For Galaxy Mobile	0.509 ms	0 - 47 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Snapdragon® X2 Elite	0.44 ms	1 - 1 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Snapdragon® X Elite	1.054 ms	0 - 0 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Snapdragon® 8 Gen 3 Mobile	0.691 ms	0 - 63 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Snapdragon® 8 Gen 1 Mobile	1.19 ms	0 - 65 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® Dragonwing™ QCS6490	2.201 ms	0 - 3 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® Dragonwing™ IQ-8275	1.062 ms	0 - 4 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.023 ms	0 - 5 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® QCS8450	1.19 ms	0 - 65 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® Dragonwing™ IQ-9075	1.187 ms	0 - 3 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® Dragonwing™ IQ-X7181	1.054 ms	0 - 0 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® Dragonwing™ Q-6690	3.609 ms	0 - 155 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® Dragonwing™ Q-7790	1.059 ms	0 - 47 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Qualcomm® Dragonwing™ Q-8750	0.509 ms	0 - 47 MB	NPU
Movenet	ONNX	w8a16_mixed_int16	Snapdragon® 7 Gen 4 Mobile	1.059 ms	0 - 47 MB	NPU
Movenet	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.416 ms	0 - 34 MB	NPU
Movenet	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	0.521 ms	0 - 38 MB	NPU
Movenet	QNN_DLC	float	Snapdragon® X2 Elite	0.627 ms	0 - 0 MB	NPU
Movenet	QNN_DLC	float	Snapdragon® X Elite	1.137 ms	0 - 0 MB	NPU
Movenet	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	0.701 ms	0 - 52 MB	NPU
Movenet	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	1.154 ms	0 - 56 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	1.198 ms	0 - 4 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.949 ms	0 - 2 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® SA8650P	1.342 ms	0 - 32 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® SA8255P	1.342 ms	0 - 32 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® QCS8450	1.154 ms	0 - 56 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	1.228 ms	0 - 3 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	1.137 ms	0 - 0 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	0.521 ms	0 - 38 MB	NPU
Movenet	QNN_DLC	float	Qualcomm® SA8295P	1.486 ms	0 - 33 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.417 ms	0 - 41 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Snapdragon® 8 Elite For Galaxy Mobile	0.483 ms	0 - 43 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Snapdragon® X2 Elite	0.636 ms	0 - 0 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Snapdragon® X Elite	1.197 ms	0 - 0 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Snapdragon® 8 Gen 3 Mobile	0.668 ms	0 - 53 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® Dragonwing™ IQ-8275	0.986 ms	0 - 3 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.02 ms	0 - 2 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® SA8775P	1.298 ms	0 - 43 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® SA8650P	1.298 ms	0 - 43 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® SA8255P	1.298 ms	0 - 43 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® Dragonwing™ IQ-9075	1.158 ms	0 - 3 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® Dragonwing™ IQ-X7181	1.197 ms	0 - 0 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® Dragonwing™ Q-6690	3.776 ms	0 - 147 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® Dragonwing™ Q-7790	1.041 ms	0 - 39 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® Dragonwing™ Q-8750	0.483 ms	0 - 43 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Qualcomm® SA7255P	2.066 ms	0 - 40 MB	NPU
Movenet	QNN_DLC	w8a16_mixed_int16	Snapdragon® 7 Gen 4 Mobile	1.041 ms	0 - 39 MB	NPU
Movenet	TFLITE	float	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.419 ms	0 - 38 MB	NPU
Movenet	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	0.524 ms	0 - 36 MB	NPU
Movenet	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	0.706 ms	0 - 55 MB	NPU
Movenet	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	1.152 ms	0 - 59 MB	NPU
Movenet	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	1.205 ms	0 - 9 MB	NPU
Movenet	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	0.946 ms	0 - 3 MB	NPU
Movenet	TFLITE	float	Qualcomm® SA8650P	1.355 ms	0 - 36 MB	NPU
Movenet	TFLITE	float	Qualcomm® SA8255P	1.355 ms	0 - 36 MB	NPU
Movenet	TFLITE	float	Qualcomm® QCS8450	1.152 ms	0 - 59 MB	NPU
Movenet	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	1.226 ms	0 - 9 MB	NPU
Movenet	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	0.524 ms	0 - 36 MB	NPU
Movenet	TFLITE	float	Qualcomm® SA7255P	2.94 ms	0 - 38 MB	NPU
Movenet	TFLITE	float	Qualcomm® SA8295P	1.468 ms	0 - 35 MB	NPU

License

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

References

• MoveNet: Ultra fast and accurate pose detection model
• 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.