RF-DETR: Optimized for Qualcomm Devices

DETR is a machine learning model that can detect objects (trained on COCO dataset).

This is based on the implementation of RF-DETR 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 2.42, ONNX Runtime 1.25.0	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

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

Model Details

Model Type: Model_use_case.object_detection

Model Stats:

• Model checkpoint: RF-DETR-base
• Input resolution: 560x560
• Number of parameters: 29.0M
• Model size: 116MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
RF-DETR	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	30.605 ms	10 - 714 MB	NPU
RF-DETR	ONNX	float	Snapdragon® 8 Gen 3 Mobile	49.507 ms	16 - 788 MB	NPU
RF-DETR	ONNX	float	Qualcomm® QCS8550 (Proxy)	67.771 ms	14 - 18 MB	NPU
RF-DETR	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	37.101 ms	13 - 618 MB	NPU
RF-DETR	ONNX	float	Qualcomm® QCS9075	91.499 ms	14 - 59 MB	NPU
RF-DETR	ONNX	float	Qualcomm® QCS8750	37.101 ms	13 - 618 MB	NPU
RF-DETR	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	31.111 ms	4 - 645 MB	NPU
RF-DETR	QNN_DLC	float	Snapdragon® X2 Elite	28.694 ms	4 - 4 MB	NPU
RF-DETR	QNN_DLC	float	Snapdragon® X Elite	66.114 ms	4 - 4 MB	NPU
RF-DETR	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	48.358 ms	4 - 765 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® QCS8275	151.087 ms	2 - 622 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	67.066 ms	4 - 33 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® SA8775P	70.585 ms	2 - 615 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® SA8650P	70.585 ms	2 - 615 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® SA8255P	70.585 ms	2 - 615 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	104.002 ms	4 - 684 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® SA7255P	151.087 ms	2 - 622 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® SA8295P	91.982 ms	0 - 609 MB	NPU
RF-DETR	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	35.116 ms	0 - 733 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® QCS9075	79.862 ms	4 - 9 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® QCS8750	35.116 ms	0 - 733 MB	NPU
RF-DETR	QNN_DLC	float	Qualcomm® QCS7181	66.114 ms	4 - 4 MB	NPU
RF-DETR	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	36.835 ms	6 - 618 MB	NPU
RF-DETR	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	57.453 ms	5 - 760 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® QCS8275	172.406 ms	7 - 611 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® QCS8550 (Proxy)	80.56 ms	7 - 11 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® SA8775P	85.706 ms	7 - 616 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® SA8650P	85.706 ms	7 - 616 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® SA8255P	85.706 ms	7 - 616 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® QCS8450 (Proxy)	118.884 ms	7 - 657 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® SA7255P	172.406 ms	7 - 611 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® SA8295P	105.551 ms	0 - 589 MB	NPU
RF-DETR	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	42.913 ms	5 - 605 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® QCS9075	92.014 ms	6 - 77 MB	NPU
RF-DETR	TFLITE	float	Qualcomm® QCS8750	42.913 ms	5 - 605 MB	NPU

License

• The license for the original implementation of RF-DETR can be found here.

References

• RF-DETR A SOTA Real-Time Object Detection Model
• 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.