DETR-ResNet50: 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 DETR-ResNet50 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 1.25.0	Download
ONNX	w8a16_mixed_fp16	Universal	QAIRT 2.45, 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 DETR-ResNet50 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 DETR-ResNet50 on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.object_detection

Model Stats:

• Model checkpoint: ResNet50
• Input resolution: 480x480
• Number of parameters: 41.4M
• Model size (float): 158 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
DETR-ResNet50	ONNX	float	Snapdragon® X2 Elite	8.217 ms	176 - 176 MB	NPU
DETR-ResNet50	ONNX	float	Snapdragon® X Elite	17.83 ms	144 - 144 MB	NPU
DETR-ResNet50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	13.328 ms	3 - 363 MB	NPU
DETR-ResNet50	ONNX	float	Snapdragon® 8 Gen 1 Mobile	36.104 ms	5 - 313 MB	NPU
DETR-ResNet50	ONNX	float	Qualcomm® QCS8550 (Proxy)	17.902 ms	0 - 142 MB	NPU
DETR-ResNet50	ONNX	float	Qualcomm® QCS8450	36.104 ms	5 - 313 MB	NPU
DETR-ResNet50	ONNX	float	Snapdragon® 8 Elite Mobile	9.951 ms	2 - 271 MB	NPU
DETR-ResNet50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	7.374 ms	3 - 311 MB	NPU
DETR-ResNet50	ONNX	float	Qualcomm® QCS9075	27.505 ms	5 - 50 MB	NPU
DETR-ResNet50	ONNX	float	Qualcomm® QCS8750	9.951 ms	2 - 271 MB	NPU
DETR-ResNet50	ONNX	float	Qualcomm® QCS7181	17.83 ms	144 - 144 MB	NPU
DETR-ResNet50	ONNX	w8a16_mixed_fp16	Snapdragon® 8 Gen 3 Mobile	52.062 ms	88 - 502 MB	NPU
DETR-ResNet50	ONNX	w8a16_mixed_fp16	Qualcomm® QCS8550 (Proxy)	67.981 ms	0 - 60 MB	NPU
DETR-ResNet50	ONNX	w8a16_mixed_fp16	Snapdragon® 8 Elite Mobile	41.988 ms	20 - 368 MB	NPU
DETR-ResNet50	ONNX	w8a16_mixed_fp16	Snapdragon® 8 Elite Gen 5 Mobile	33.05 ms	21 - 387 MB	NPU
DETR-ResNet50	ONNX	w8a16_mixed_fp16	Qualcomm® QCS9075	101.433 ms	22 - 65 MB	NPU
DETR-ResNet50	ONNX	w8a16_mixed_fp16	Qualcomm® QCS8750	41.988 ms	20 - 368 MB	NPU
DETR-ResNet50	QNN_DLC	float	Snapdragon® X2 Elite	9.014 ms	5 - 5 MB	NPU
DETR-ResNet50	QNN_DLC	float	Snapdragon® X Elite	19.532 ms	5 - 5 MB	NPU
DETR-ResNet50	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	13.634 ms	0 - 351 MB	NPU
DETR-ResNet50	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	42.016 ms	5 - 304 MB	NPU
DETR-ResNet50	QNN_DLC	float	Qualcomm® QCS8275	86.004 ms	1 - 268 MB	NPU
DETR-ResNet50	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	19.308 ms	5 - 12 MB	NPU
DETR-ResNet50	QNN_DLC	float	Qualcomm® QCS8450	42.016 ms	5 - 304 MB	NPU
DETR-ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Mobile	10.436 ms	0 - 273 MB	NPU
DETR-ResNet50	QNN_DLC	float	Qualcomm® SA8295P	31.193 ms	0 - 221 MB	NPU
DETR-ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	7.84 ms	5 - 289 MB	NPU
DETR-ResNet50	QNN_DLC	float	Qualcomm® SA7255P	86.004 ms	1 - 268 MB	NPU
DETR-ResNet50	QNN_DLC	float	Qualcomm® QCS9075	31.038 ms	5 - 11 MB	NPU
DETR-ResNet50	QNN_DLC	float	Qualcomm® QCS8750	10.436 ms	0 - 273 MB	NPU
DETR-ResNet50	QNN_DLC	float	Qualcomm® QCS7181	19.532 ms	5 - 5 MB	NPU
DETR-ResNet50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	13.625 ms	0 - 386 MB	NPU
DETR-ResNet50	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	41.35 ms	0 - 329 MB	NPU
DETR-ResNet50	TFLITE	float	Qualcomm® QCS8275	86.031 ms	0 - 293 MB	NPU
DETR-ResNet50	TFLITE	float	Qualcomm® QCS8550 (Proxy)	19.121 ms	0 - 3 MB	NPU
DETR-ResNet50	TFLITE	float	Qualcomm® SA8775P	601.937 ms	0 - 12 MB	CPU
DETR-ResNet50	TFLITE	float	Qualcomm® SA8650P	601.937 ms	0 - 12 MB	CPU
DETR-ResNet50	TFLITE	float	Qualcomm® SA8255P	601.937 ms	0 - 12 MB	CPU
DETR-ResNet50	TFLITE	float	Qualcomm® QCS8450	41.35 ms	0 - 329 MB	NPU
DETR-ResNet50	TFLITE	float	Snapdragon® 8 Elite Mobile	10.52 ms	0 - 306 MB	NPU
DETR-ResNet50	TFLITE	float	Qualcomm® SA8295P	31.09 ms	0 - 247 MB	NPU
DETR-ResNet50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	7.993 ms	0 - 310 MB	NPU
DETR-ResNet50	TFLITE	float	Qualcomm® SA7255P	86.031 ms	0 - 293 MB	NPU
DETR-ResNet50	TFLITE	float	Qualcomm® QCS9075	30.4 ms	0 - 88 MB	NPU
DETR-ResNet50	TFLITE	float	Qualcomm® QCS8750	10.52 ms	0 - 306 MB	NPU

License

• The license for the original implementation of DETR-ResNet50 can be found here.

References

• End-to-End Object Detection with Transformers
• 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.