Modelpublic

qualcomm/Yolo-X

sourceHugging Faceapache-2.0updated 12d agoView on Hugging Face

9likes959downloads

Model Card

Yolo-X: Optimized for Qualcomm Devices

YoloX is a machine learning model that predicts bounding boxes and classes of objects in an image.

This is based on the implementation of Yolo-X 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 2.45, ONNX Runtime 1.27.1	Download
ONNX	w8a16	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	Download
ONNX	w8a8	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	Download
ONNX	w8a8mixedint16	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	Download
QNN_DLC	w8a16	Universal	QAIRT 2.45	Download
QNN_DLC	w8a8	Universal	QAIRT 2.45	Download
TFLITE	w8a8	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit [Yolo-X on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/yolox).

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 Yolo-X on GitHub for usage instructions.

Model Details

Model Type: Modelusecase.object_detection

Model Stats:

—Input resolution: 640x640
—Model checkpoint: YoloX Small
—Model size (float): 34.3 MB
—Model size (w8a16): 9.53 MB
—Model size (w8a8): 8.96 MB
—Number of parameters: 8.98M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Yolo-X	ONNX	float	Snapdragon® X2 Elite	7.683 ms	5 - 5 MB	NPU
Yolo-X	ONNX	float	Snapdragon® X Elite	13.812 ms	14 - 14 MB	NPU
Yolo-X	ONNX	float	Snapdragon® 8 Gen 3 Mobile	9.254 ms	1 - 184 MB	NPU
Yolo-X	ONNX	float	Snapdragon® 8 Gen 1 Mobile	25.917 ms	1 - 209 MB	NPU
Yolo-X	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	15.417 ms	5 - 8 MB	NPU
Yolo-X	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	13.606 ms	0 - 143 MB	NPU
Yolo-X	ONNX	float	Qualcomm® QCS8450	25.917 ms	1 - 209 MB	NPU
Yolo-X	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	16.616 ms	5 - 7 MB	NPU
Yolo-X	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	13.812 ms	14 - 14 MB	NPU
Yolo-X	ONNX	float	Qualcomm® Dragonwing™ Q-8750	7.728 ms	1 - 160 MB	NPU
Yolo-X	ONNX	float	Snapdragon® 8 Elite Mobile	7.728 ms	1 - 160 MB	NPU
Yolo-X	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	6.261 ms	2 - 165 MB	NPU
Yolo-X	ONNX	w8a16	Snapdragon® X2 Elite	11.388 ms	2 - 2 MB	NPU
Yolo-X	ONNX	w8a16	Snapdragon® X Elite	15.074 ms	8 - 8 MB	NPU
Yolo-X	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	9.995 ms	5 - 242 MB	NPU
Yolo-X	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	16.206 ms	3 - 241 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® Dragonwing™ QCS6490	29.176 ms	2 - 5 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-8275	13.741 ms	2 - 6 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	14.114 ms	0 - 10 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® QCS8450	16.206 ms	3 - 241 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-9075	15.504 ms	2 - 5 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-X7181	15.074 ms	8 - 8 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® Dragonwing™ Q-6690	67.427 ms	3 - 206 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® Dragonwing™ Q-7790	10.613 ms	3 - 206 MB	NPU
Yolo-X	ONNX	w8a16	Qualcomm® Dragonwing™ Q-8750	8.897 ms	1 - 204 MB	NPU
Yolo-X	ONNX	w8a16	Snapdragon® 8 Elite Mobile	8.897 ms	1 - 204 MB	NPU
Yolo-X	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	11.076 ms	0 - 208 MB	NPU
Yolo-X	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	10.613 ms	3 - 206 MB	NPU
Yolo-X	ONNX	w8a8	Snapdragon® X2 Elite	8.239 ms	1 - 1 MB	NPU
Yolo-X	ONNX	w8a8	Snapdragon® X Elite	8.95 ms	8 - 8 MB	NPU
Yolo-X	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	6.121 ms	1 - 90 MB	NPU
Yolo-X	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	8.875 ms	1 - 91 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® Dragonwing™ QCS6490	13.544 ms	1 - 4 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-8275	8.273 ms	1 - 5 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	8.283 ms	0 - 11 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® QCS8450	8.875 ms	1 - 91 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-9075	8.44 ms	1 - 4 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-X7181	8.95 ms	8 - 8 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® Dragonwing™ Q-6690	29.433 ms	1 - 176 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® Dragonwing™ Q-7790	8.952 ms	1 - 176 MB	NPU
Yolo-X	ONNX	w8a8	Qualcomm® Dragonwing™ Q-8750	5.698 ms	1 - 60 MB	NPU
Yolo-X	ONNX	w8a8	Snapdragon® 8 Elite Mobile	5.698 ms	1 - 60 MB	NPU
Yolo-X	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	8.092 ms	0 - 59 MB	NPU
Yolo-X	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	8.952 ms	1 - 176 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Snapdragon® X2 Elite	28.469 ms	16 - 16 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Snapdragon® X Elite	35.094 ms	19 - 19 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Snapdragon® 8 Gen 3 Mobile	26.62 ms	11 - 858 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Qualcomm® Dragonwing™ IQ-8275	34.799 ms	9 - 12 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	34.609 ms	4 - 28 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Qualcomm® Dragonwing™ IQ-9075	35.639 ms	11 - 14 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Qualcomm® Dragonwing™ IQ-X7181	35.094 ms	19 - 19 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Qualcomm® Dragonwing™ Q-8750	23.481 ms	11 - 804 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Snapdragon® 8 Elite Mobile	23.481 ms	11 - 804 MB	NPU
Yolo-X	ONNX	w8a8mixedint16	Snapdragon® 8 Elite Gen 5 Mobile	25.81 ms	11 - 839 MB	NPU
Yolo-X	QNN_DLC	w8a16	Snapdragon® X2 Elite	2.817 ms	2 - 2 MB	NPU
Yolo-X	QNN_DLC	w8a16	Snapdragon® X Elite	6.034 ms	2 - 2 MB	NPU
Yolo-X	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	3.663 ms	2 - 105 MB	NPU
Yolo-X	QNN_DLC	w8a16	Snapdragon® 8 Gen 1 Mobile	6.867 ms	2 - 107 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS6490	22.816 ms	2 - 6 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-8275	5.246 ms	2 - 6 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	5.528 ms	2 - 132 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® SA8775P	6.171 ms	0 - 73 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® SA8650P	6.171 ms	0 - 73 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® SA8255P	6.171 ms	0 - 73 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® QCS8450	6.867 ms	2 - 107 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	5.868 ms	0 - 4 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	6.034 ms	2 - 2 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-6690	48.142 ms	2 - 195 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-7790	6.974 ms	2 - 190 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	2.598 ms	2 - 80 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® SA7255P	11.67 ms	1 - 72 MB	NPU
Yolo-X	QNN_DLC	w8a16	Qualcomm® SA8295P	7.581 ms	0 - 69 MB	NPU
Yolo-X	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	2.598 ms	2 - 80 MB	NPU
Yolo-X	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	1.965 ms	2 - 85 MB	NPU
Yolo-X	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	6.974 ms	2 - 190 MB	NPU
Yolo-X	QNN_DLC	w8a8	Snapdragon® X2 Elite	1.226 ms	1 - 1 MB	NPU
Yolo-X	QNN_DLC	w8a8	Snapdragon® X Elite	2.739 ms	1 - 1 MB	NPU
Yolo-X	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	1.654 ms	1 - 75 MB	NPU
Yolo-X	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	3.001 ms	1 - 76 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS6490	8.763 ms	1 - 3 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-8275	2.416 ms	1 - 5 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	2.446 ms	1 - 2 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® SA8775P	2.805 ms	1 - 53 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® SA8650P	2.805 ms	1 - 53 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® SA8255P	2.805 ms	1 - 53 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® QCS8450	3.001 ms	1 - 76 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-9075	2.544 ms	3 - 6 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-X7181	2.739 ms	1 - 1 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-6690	23.958 ms	16 - 182 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-7790	2.784 ms	1 - 49 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-8750	1.171 ms	0 - 46 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® SA7255P	5.86 ms	1 - 49 MB	NPU
Yolo-X	QNN_DLC	w8a8	Qualcomm® SA8295P	3.798 ms	1 - 48 MB	NPU
Yolo-X	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	1.171 ms	0 - 46 MB	NPU
Yolo-X	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.91 ms	1 - 53 MB	NPU
Yolo-X	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	2.784 ms	1 - 49 MB	NPU
Yolo-X	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	1.649 ms	0 - 78 MB	NPU
Yolo-X	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	2.977 ms	0 - 75 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® Dragonwing™ QCS6490	8.846 ms	0 - 12 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-8275	2.436 ms	0 - 13 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	2.455 ms	0 - 5 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® SA8775P	2.916 ms	0 - 51 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® SA8650P	2.916 ms	0 - 51 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® SA8255P	2.916 ms	0 - 51 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® QCS8450	2.977 ms	0 - 75 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-9075	2.649 ms	0 - 12 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-6690	21.686 ms	0 - 158 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-7790	2.769 ms	0 - 45 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-8750	1.194 ms	0 - 44 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® SA7255P	5.894 ms	0 - 46 MB	NPU
Yolo-X	TFLITE	w8a8	Qualcomm® SA8295P	3.815 ms	0 - 45 MB	NPU
Yolo-X	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	1.194 ms	0 - 44 MB	NPU
Yolo-X	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.92 ms	0 - 51 MB	NPU
Yolo-X	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	2.769 ms	0 - 45 MB	NPU

License

—The license for the original implementation of Yolo-X can be found here.

References

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.