RTMPose-Body2d: Optimized for Qualcomm Devices

RTMPose is a machine learning model that detects human pose and returns a location and confidence for each of 133 joints.

This is based on the implementation of RTMPose-Body2d 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 1.30.0	Download
ONNX	w8a16	Universal	QAIRT 2.50, ONNX Runtime 1.30.0	Download
QNN_DLC	float	Universal	QAIRT 2.50	Download
TFLITE	float	Universal	QAIRT 2.50	Download

For more device-specific assets and performance metrics, visit RTMPose-Body2d 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 RTMPose-Body2d on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.pose_estimation

Model Stats:

• Input resolution: 256x192
• Model size (float): 68.5 MB
• Model size (w8a16): 18.2 MB
• Number of parameters: 17.9M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
RTMPose-Body2d	ONNX	float	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.942 ms	0 - 43 MB	NPU
RTMPose-Body2d	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1.09 ms	0 - 43 MB	NPU
RTMPose-Body2d	ONNX	float	Snapdragon® X2 Elite	0.918 ms	2 - 2 MB	NPU
RTMPose-Body2d	ONNX	float	Snapdragon® X Elite	1.724 ms	37 - 37 MB	NPU
RTMPose-Body2d	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.306 ms	0 - 58 MB	NPU
RTMPose-Body2d	ONNX	float	Snapdragon® 8 Gen 1 Mobile	3.191 ms	0 - 68 MB	NPU
RTMPose-Body2d	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	2.584 ms	1 - 6 MB	NPU
RTMPose-Body2d	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.773 ms	0 - 42 MB	NPU
RTMPose-Body2d	ONNX	float	Qualcomm® QCS8450	3.191 ms	0 - 68 MB	NPU
RTMPose-Body2d	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	2.324 ms	0 - 4 MB	NPU
RTMPose-Body2d	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	1.724 ms	37 - 37 MB	NPU
RTMPose-Body2d	ONNX	float	Qualcomm® Dragonwing™ Q-8750	1.09 ms	0 - 43 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.761 ms	0 - 65 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	0.874 ms	0 - 62 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Snapdragon® X2 Elite	0.788 ms	1 - 1 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Snapdragon® X Elite	1.819 ms	19 - 19 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	1.213 ms	0 - 95 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	2.192 ms	0 - 94 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® Dragonwing™ QCS6490	6.364 ms	2 - 5 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-8275	1.755 ms	0 - 4 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.783 ms	0 - 22 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® QCS8450	2.192 ms	0 - 94 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-9075	1.921 ms	0 - 3 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-X7181	1.819 ms	19 - 19 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® Dragonwing™ Q-6690	12.676 ms	2 - 173 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® Dragonwing™ Q-7790	2.455 ms	0 - 167 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Qualcomm® Dragonwing™ Q-8750	0.874 ms	0 - 62 MB	NPU
RTMPose-Body2d	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	2.455 ms	0 - 167 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.934 ms	1 - 39 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.067 ms	1 - 36 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Snapdragon® X2 Elite	1.093 ms	1 - 1 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Snapdragon® X Elite	1.878 ms	1 - 1 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.297 ms	0 - 54 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	3.143 ms	0 - 66 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	2.52 ms	1 - 5 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.744 ms	1 - 2 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® SA8775P	2.41 ms	1 - 39 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® SA8650P	2.41 ms	1 - 39 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® SA8255P	2.41 ms	1 - 39 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® QCS8450	3.143 ms	0 - 66 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	2.294 ms	1 - 4 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	1.878 ms	1 - 1 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	1.067 ms	1 - 36 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® SA7255P	7.552 ms	1 - 35 MB	NPU
RTMPose-Body2d	QNN_DLC	float	Qualcomm® SA8295P	3.407 ms	0 - 38 MB	NPU
RTMPose-Body2d	TFLITE	float	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.998 ms	0 - 49 MB	NPU
RTMPose-Body2d	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1.13 ms	0 - 48 MB	NPU
RTMPose-Body2d	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.36 ms	0 - 75 MB	NPU
RTMPose-Body2d	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	3.299 ms	0 - 87 MB	NPU
RTMPose-Body2d	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	2.863 ms	1 - 43 MB	NPU
RTMPose-Body2d	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.817 ms	1 - 4 MB	NPU
RTMPose-Body2d	TFLITE	float	Qualcomm® QCS8450	3.299 ms	0 - 87 MB	NPU
RTMPose-Body2d	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	2.505 ms	1 - 42 MB	NPU
RTMPose-Body2d	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	1.13 ms	0 - 48 MB	NPU
RTMPose-Body2d	TFLITE	float	Qualcomm® SA7255P	8.268 ms	1 - 50 MB	NPU
RTMPose-Body2d	TFLITE	float	Qualcomm® SA8295P	3.76 ms	1 - 51 MB	NPU

License

• The license for the original implementation of RTMPose-Body2d can be found here.

References

• RTMPose: Real-Time Multi-Person Pose Estimation based on MMPose
• 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.