Yolo-v3: Optimized for Qualcomm Devices

YoloV3 is a machine learning model that predicts bounding boxes and classes of objects in an image.

This is based on the implementation of Yolo-v3 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-v3 on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.object_detection

Model Stats:

• Model checkpoint: YoloV3 Tiny
• Input resolution: 416p (416x416)
• Number of parameters: 11.5M
• Model size (float): 43.9 MB
• Model size (w8a16): 16.9 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Yolo-v3	ONNX	float	Snapdragon® X2 Elite	2.432 ms	208 - 208 MB	NPU
Yolo-v3	ONNX	float	Snapdragon® X Elite	4.072 ms	144 - 144 MB	NPU
Yolo-v3	ONNX	float	Snapdragon® 8 Gen 3 Mobile	3.109 ms	2 - 156 MB	NPU
Yolo-v3	ONNX	float	Snapdragon® 8 Gen 1 Mobile	10.499 ms	5 - 183 MB	NPU
Yolo-v3	ONNX	float	Qualcomm® QCS8550 (Proxy)	3.979 ms	0 - 26 MB	NPU
Yolo-v3	ONNX	float	Qualcomm® QCS8450	10.499 ms	5 - 183 MB	NPU
Yolo-v3	ONNX	float	Snapdragon® 8 Elite Mobile	2.548 ms	3 - 152 MB	NPU
Yolo-v3	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.839 ms	0 - 147 MB	NPU
Yolo-v3	ONNX	float	Qualcomm® QCS9075	6.215 ms	5 - 50 MB	NPU
Yolo-v3	ONNX	float	Qualcomm® QCS8750	2.548 ms	3 - 152 MB	NPU
Yolo-v3	ONNX	float	Qualcomm® QCS7181	4.072 ms	144 - 144 MB	NPU
Yolo-v3	ONNX	w8a16	Snapdragon® X2 Elite	1.601 ms	210 - 210 MB	NPU
Yolo-v3	ONNX	w8a16	Snapdragon® X Elite	3.824 ms	147 - 147 MB	NPU
Yolo-v3	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	2.516 ms	2 - 79 MB	NPU
Yolo-v3	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	4.531 ms	2 - 79 MB	NPU
Yolo-v3	ONNX	w8a16	Qualcomm® QCS6490	12.244 ms	2 - 48 MB	NPU
Yolo-v3	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	3.657 ms	0 - 13 MB	NPU
Yolo-v3	ONNX	w8a16	Qualcomm® QCS8450	4.531 ms	2 - 79 MB	NPU
Yolo-v3	ONNX	w8a16	Qualcomm® QCS9075	3.981 ms	2 - 48 MB	NPU
Yolo-v3	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	1.376 ms	0 - 52 MB	NPU
Yolo-v3	ONNX	w8a16	Snapdragon® 8 Elite Mobile	1.879 ms	1 - 55 MB	NPU
Yolo-v3	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	5.225 ms	2 - 176 MB	NPU
Yolo-v3	ONNX	w8a16	Qualcomm® QCM6690	39.679 ms	3 - 182 MB	NPU
Yolo-v3	ONNX	w8a16	Qualcomm® QCS7790	5.225 ms	2 - 176 MB	NPU
Yolo-v3	ONNX	w8a16	Qualcomm® QCS8750	1.879 ms	1 - 55 MB	NPU
Yolo-v3	ONNX	w8a16	Qualcomm® QCS7181	3.824 ms	147 - 147 MB	NPU
Yolo-v3	QNN_DLC	float	Snapdragon® X2 Elite	2.48 ms	5 - 5 MB	NPU
Yolo-v3	QNN_DLC	float	Snapdragon® X Elite	3.672 ms	5 - 5 MB	NPU
Yolo-v3	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.742 ms	0 - 161 MB	NPU
Yolo-v3	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	10.045 ms	5 - 174 MB	NPU
Yolo-v3	QNN_DLC	float	Qualcomm® QCS8275	20.113 ms	0 - 143 MB	NPU
Yolo-v3	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	3.412 ms	5 - 15 MB	NPU
Yolo-v3	QNN_DLC	float	Qualcomm® QCS8450	10.045 ms	5 - 174 MB	NPU
Yolo-v3	QNN_DLC	float	Snapdragon® 8 Elite Mobile	2.153 ms	0 - 142 MB	NPU
Yolo-v3	QNN_DLC	float	Qualcomm® SA8295P	9.176 ms	0 - 150 MB	NPU
Yolo-v3	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.671 ms	5 - 155 MB	NPU
Yolo-v3	QNN_DLC	float	Qualcomm® SA7255P	20.113 ms	0 - 143 MB	NPU
Yolo-v3	QNN_DLC	float	Qualcomm® QCS9075	5.802 ms	7 - 13 MB	NPU
Yolo-v3	QNN_DLC	float	Qualcomm® QCS8750	2.153 ms	0 - 142 MB	NPU
Yolo-v3	QNN_DLC	float	Qualcomm® QCS7181	3.672 ms	5 - 5 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Snapdragon® X2 Elite	2.182 ms	2 - 2 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Snapdragon® X Elite	4.025 ms	2 - 2 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	2.57 ms	2 - 75 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Snapdragon® 8 Gen 1 Mobile	4.387 ms	2 - 82 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCS6490	11.875 ms	4 - 8 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCS8275	9.217 ms	2 - 47 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCS8550 (Proxy)	3.666 ms	2 - 4 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCS8450	4.387 ms	2 - 82 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCS9075	3.942 ms	3 - 7 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	1.399 ms	2 - 56 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	1.886 ms	2 - 54 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® SA8295P	5.488 ms	0 - 43 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	4.641 ms	2 - 173 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCM6690	38.181 ms	2 - 175 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® SA7255P	9.217 ms	2 - 47 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCS7790	4.641 ms	2 - 173 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCS8750	1.886 ms	2 - 54 MB	NPU
Yolo-v3	QNN_DLC	w8a16	Qualcomm® QCS7181	4.025 ms	2 - 2 MB	NPU
Yolo-v3	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.678 ms	0 - 188 MB	NPU
Yolo-v3	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	10.011 ms	0 - 195 MB	NPU
Yolo-v3	TFLITE	float	Qualcomm® QCS8275	19.946 ms	0 - 145 MB	NPU
Yolo-v3	TFLITE	float	Qualcomm® QCS8550 (Proxy)	3.329 ms	0 - 2 MB	NPU
Yolo-v3	TFLITE	float	Qualcomm® SA8775P	32.628 ms	1 - 24 MB	GPU
Yolo-v3	TFLITE	float	Qualcomm® SA8650P	32.628 ms	1 - 24 MB	GPU
Yolo-v3	TFLITE	float	Qualcomm® SA8255P	32.628 ms	1 - 24 MB	GPU
Yolo-v3	TFLITE	float	Qualcomm® QCS8450	10.011 ms	0 - 195 MB	NPU
Yolo-v3	TFLITE	float	Snapdragon® 8 Elite Mobile	2.123 ms	0 - 144 MB	NPU
Yolo-v3	TFLITE	float	Qualcomm® SA8295P	9.103 ms	0 - 152 MB	NPU
Yolo-v3	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.54 ms	0 - 150 MB	NPU
Yolo-v3	TFLITE	float	Qualcomm® SA7255P	19.946 ms	0 - 145 MB	NPU
Yolo-v3	TFLITE	float	Qualcomm® QCS9075	5.578 ms	0 - 30 MB	NPU
Yolo-v3	TFLITE	float	Qualcomm® QCS8750	2.123 ms	0 - 144 MB	NPU

License

• The license for the original implementation of Yolo-v3 can be found here.

References

• YOLOv3: An Incremental Improvement
• 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.