Yolo-R: Optimized for Qualcomm Devices

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

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

Model Details

Model Type: Model_use_case.object_detection

Model Stats:

• Model checkpoint: yolor_p6
• Input resolution: 640x640
• Number of parameters: 4.68M
• Model size (float): 17.9 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Yolo-R	ONNX	float	Snapdragon® X2 Elite	26.309 ms	75 - 75 MB	NPU
Yolo-R	ONNX	float	Snapdragon® X Elite	57.577 ms	74 - 74 MB	NPU
Yolo-R	ONNX	float	Snapdragon® 8 Gen 3 Mobile	41.986 ms	2 - 346 MB	NPU
Yolo-R	ONNX	float	Qualcomm® QCS8550 (Proxy)	55.998 ms	0 - 79 MB	NPU
Yolo-R	ONNX	float	Qualcomm® QCS9075	52.751 ms	5 - 12 MB	NPU
Yolo-R	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	37.052 ms	3 - 233 MB	NPU
Yolo-R	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	26.443 ms	6 - 318 MB	NPU
Yolo-R	ONNX	w8a16	Snapdragon® X2 Elite	18.322 ms	41 - 41 MB	NPU
Yolo-R	ONNX	w8a16	Snapdragon® X Elite	30.49 ms	40 - 40 MB	NPU
Yolo-R	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	20.897 ms	0 - 490 MB	NPU
Yolo-R	ONNX	w8a16	Qualcomm® QCS6490	2282.662 ms	129 - 136 MB	CPU
Yolo-R	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	29.166 ms	0 - 50 MB	NPU
Yolo-R	ONNX	w8a16	Qualcomm® QCS9075	29.553 ms	1 - 6 MB	NPU
Yolo-R	ONNX	w8a16	Qualcomm® QCM6690	1176.534 ms	48 - 62 MB	CPU
Yolo-R	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	17.119 ms	1 - 364 MB	NPU
Yolo-R	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	1118.647 ms	154 - 166 MB	CPU
Yolo-R	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	17.934 ms	3 - 429 MB	NPU
Yolo-R	QNN_DLC	float	Snapdragon® X2 Elite	15.23 ms	5 - 5 MB	NPU
Yolo-R	QNN_DLC	float	Snapdragon® X Elite	29.418 ms	5 - 5 MB	NPU
Yolo-R	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	22.761 ms	0 - 317 MB	NPU
Yolo-R	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	100.329 ms	0 - 252 MB	NPU
Yolo-R	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	29.567 ms	5 - 7 MB	NPU
Yolo-R	QNN_DLC	float	Qualcomm® SA8775P	36.109 ms	0 - 259 MB	NPU
Yolo-R	QNN_DLC	float	Qualcomm® QCS9075	37.531 ms	5 - 11 MB	NPU
Yolo-R	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	50.546 ms	5 - 377 MB	NPU
Yolo-R	QNN_DLC	float	Qualcomm® SA7255P	100.329 ms	0 - 252 MB	NPU
Yolo-R	QNN_DLC	float	Qualcomm® SA8295P	44.017 ms	0 - 307 MB	NPU
Yolo-R	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	18.262 ms	0 - 230 MB	NPU
Yolo-R	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	13.999 ms	5 - 295 MB	NPU
Yolo-R	QNN_DLC	w8a16	Snapdragon® X2 Elite	8.363 ms	2 - 2 MB	NPU
Yolo-R	QNN_DLC	w8a16	Snapdragon® X Elite	18.967 ms	2 - 2 MB	NPU
Yolo-R	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	12.323 ms	2 - 370 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® QCS6490	77.324 ms	3 - 7 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® QCS8275 (Proxy)	37.394 ms	0 - 288 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® QCS8550 (Proxy)	18.116 ms	2 - 4 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® SA8775P	18.34 ms	0 - 288 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® QCS9075	20.195 ms	1 - 5 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® QCM6690	246.253 ms	2 - 400 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® QCS8450 (Proxy)	25.017 ms	2 - 371 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® SA7255P	37.394 ms	0 - 288 MB	NPU
Yolo-R	QNN_DLC	w8a16	Qualcomm® SA8295P	23.495 ms	0 - 291 MB	NPU
Yolo-R	QNN_DLC	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	9.865 ms	0 - 299 MB	NPU
Yolo-R	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	26.414 ms	2 - 312 MB	NPU
Yolo-R	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	7.295 ms	2 - 303 MB	NPU
Yolo-R	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	39.801 ms	0 - 445 MB	NPU
Yolo-R	TFLITE	float	Qualcomm® QCS8275 (Proxy)	137.818 ms	1 - 328 MB	NPU
Yolo-R	TFLITE	float	Qualcomm® QCS8550 (Proxy)	59.564 ms	1 - 3 MB	NPU
Yolo-R	TFLITE	float	Qualcomm® SA8775P	66.658 ms	1 - 310 MB	NPU
Yolo-R	TFLITE	float	Qualcomm® QCS9075	50.464 ms	1 - 86 MB	NPU
Yolo-R	TFLITE	float	Qualcomm® QCS8450 (Proxy)	88.543 ms	1 - 471 MB	NPU
Yolo-R	TFLITE	float	Qualcomm® SA7255P	137.818 ms	1 - 328 MB	NPU
Yolo-R	TFLITE	float	Qualcomm® SA8295P	76.899 ms	1 - 355 MB	NPU
Yolo-R	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	31.258 ms	1 - 281 MB	NPU
Yolo-R	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	26.051 ms	0 - 341 MB	NPU

License

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

References

• You Only Learn One Representation: Unified Network for Multiple Tasks
• 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.