Yolo-v6: Optimized for Qualcomm Devices

YoloV6 is a machine learning model that predicts bounding boxes and classes of objects in an image.

This is based on the implementation of Yolo-v6 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

Due to licensing restrictions, we cannot distribute pre-exported model assets for this model. 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

See our repository for Yolo-v6 on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.object_detection

Model Stats:

• Model checkpoint: YoloV6-N
• Input resolution: 640x640
• Number of parameters: 4.68M
• Model size (float): 17.9 MB
• Model size (w8a8): 4.68 MB
• Model size (w8a16): 5.03 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Yolo-v6	ONNX	float	Snapdragon® X2 Elite	3.713 ms	14 - 14 MB	NPU
Yolo-v6	ONNX	float	Snapdragon® X Elite	8.341 ms	14 - 14 MB	NPU
Yolo-v6	ONNX	float	Snapdragon® 8 Gen 3 Mobile	5.522 ms	5 - 206 MB	NPU
Yolo-v6	ONNX	float	Qualcomm® QCS8550 (Proxy)	8.135 ms	0 - 141 MB	NPU
Yolo-v6	ONNX	float	Qualcomm® QCS9075	9.713 ms	5 - 7 MB	NPU
Yolo-v6	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	4.317 ms	0 - 166 MB	NPU
Yolo-v6	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	3.989 ms	0 - 167 MB	NPU
Yolo-v6	ONNX	w8a16	Snapdragon® X2 Elite	2.231 ms	4 - 4 MB	NPU
Yolo-v6	ONNX	w8a16	Snapdragon® X Elite	4.513 ms	3 - 3 MB	NPU
Yolo-v6	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	2.674 ms	0 - 221 MB	NPU
Yolo-v6	ONNX	w8a16	Qualcomm® QCS6490	288.525 ms	40 - 44 MB	CPU
Yolo-v6	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	4.094 ms	2 - 5 MB	NPU
Yolo-v6	ONNX	w8a16	Qualcomm® QCS9075	4.852 ms	2 - 5 MB	NPU
Yolo-v6	ONNX	w8a16	Qualcomm® QCM6690	148.157 ms	42 - 50 MB	CPU
Yolo-v6	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	2.082 ms	0 - 181 MB	NPU
Yolo-v6	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	132.442 ms	49 - 58 MB	CPU
Yolo-v6	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	1.879 ms	0 - 187 MB	NPU
Yolo-v6	QNN_DLC	float	Snapdragon® X2 Elite	3.167 ms	5 - 5 MB	NPU
Yolo-v6	QNN_DLC	float	Snapdragon® X Elite	6.269 ms	5 - 5 MB	NPU
Yolo-v6	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	4.479 ms	5 - 186 MB	NPU
Yolo-v6	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	6.119 ms	5 - 22 MB	NPU
Yolo-v6	QNN_DLC	float	Qualcomm® SA8775P	7.668 ms	1 - 159 MB	NPU
Yolo-v6	QNN_DLC	float	Qualcomm® QCS9075	7.754 ms	7 - 13 MB	NPU
Yolo-v6	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	8.913 ms	5 - 184 MB	NPU
Yolo-v6	QNN_DLC	float	Qualcomm® SA8295P	9.084 ms	2 - 155 MB	NPU
Yolo-v6	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	3.32 ms	0 - 153 MB	NPU
Yolo-v6	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	3.125 ms	5 - 161 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Snapdragon® X2 Elite	1.401 ms	2 - 2 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Snapdragon® X Elite	2.514 ms	2 - 2 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	1.481 ms	2 - 58 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® QCS6490	6.625 ms	2 - 6 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® QCS8275 (Proxy)	5.397 ms	1 - 38 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® QCS8550 (Proxy)	2.207 ms	2 - 11 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® SA8775P	2.811 ms	0 - 41 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® QCS9075	2.56 ms	3 - 7 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® QCM6690	17.468 ms	2 - 152 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® QCS8450 (Proxy)	2.816 ms	2 - 58 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® SA7255P	5.397 ms	1 - 38 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Qualcomm® SA8295P	3.49 ms	0 - 35 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	1.148 ms	2 - 44 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	2.749 ms	2 - 153 MB	NPU
Yolo-v6	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	0.949 ms	2 - 43 MB	NPU
Yolo-v6	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	12.103 ms	0 - 72 MB	GPU
Yolo-v6	TFLITE	float	Qualcomm® QCS8275 (Proxy)	67.403 ms	0 - 51 MB	GPU
Yolo-v6	TFLITE	float	Qualcomm® QCS8550 (Proxy)	15.503 ms	0 - 46 MB	GPU
Yolo-v6	TFLITE	float	Qualcomm® SA8775P	25.993 ms	0 - 55 MB	GPU
Yolo-v6	TFLITE	float	Qualcomm® QCS9075	7.892 ms	0 - 18 MB	NPU
Yolo-v6	TFLITE	float	Qualcomm® QCS8450 (Proxy)	20.031 ms	0 - 80 MB	GPU
Yolo-v6	TFLITE	float	Qualcomm® SA7255P	67.403 ms	0 - 51 MB	GPU
Yolo-v6	TFLITE	float	Qualcomm® SA8295P	19.959 ms	0 - 55 MB	GPU
Yolo-v6	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	3.766 ms	0 - 163 MB	NPU
Yolo-v6	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	3.391 ms	0 - 165 MB	NPU

License

• The license for the original implementation of Yolo-v6 can be found here.

References

• YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications
• 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.