BEVDet: Optimized for Qualcomm Devices

BEVDet is a machine learning model for generating a birds eye view represenation from the sensors(cameras) mounted on a vehicle.

This is based on the implementation of BEVDet 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	ONNX Runtime 1.23.0	Download
ONNX	w8a16_mixed_fp16	Universal	ONNX Runtime 1.23.0	Download
TFLITE	float	Universal	TFLite 2.17.0	Download

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

Model Details

Model Type: Model_use_case.driver_assistance

Model Stats:

• Model checkpoint: bevdet-r50.pth
• Input resolution: 1 x 6 x 3 x 256 x 704
• Number of parameters: 44M
• Model size: 171 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
BEVDet	ONNX	float	Snapdragon® X Elite	680.081 ms	600 - 600 MB	CPU
BEVDet	ONNX	float	Snapdragon® 8 Gen 3 Mobile	2323.529 ms	278 - 288 MB	CPU
BEVDet	ONNX	float	Qualcomm® QCS8550 (Proxy)	2565.452 ms	274 - 281 MB	CPU
BEVDet	ONNX	float	Qualcomm® QCS9075	1536.227 ms	304 - 320 MB	CPU
BEVDet	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1625.923 ms	275 - 287 MB	CPU
BEVDet	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1681.251 ms	299 - 310 MB	CPU
BEVDet	ONNX	w8a16_mixed_fp16	Snapdragon® X Elite	805.09 ms	1105 - 1105 MB	CPU
BEVDet	ONNX	w8a16_mixed_fp16	Snapdragon® 8 Gen 3 Mobile	2317.188 ms	258 - 273 MB	CPU
BEVDet	ONNX	w8a16_mixed_fp16	Qualcomm® QCS8550 (Proxy)	2662.588 ms	311 - 325 MB	CPU
BEVDet	ONNX	w8a16_mixed_fp16	Qualcomm® QCS9075	1750.335 ms	338 - 354 MB	CPU
BEVDet	ONNX	w8a16_mixed_fp16	Snapdragon® 8 Elite For Galaxy Mobile	1711.346 ms	319 - 329 MB	CPU
BEVDet	ONNX	w8a16_mixed_fp16	Snapdragon® 8 Elite Gen 5 Mobile	1831.786 ms	314 - 327 MB	CPU
BEVDet	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1670.163 ms	123 - 139 MB	CPU
BEVDet	TFLITE	float	Qualcomm® QCS8275 (Proxy)	3149.836 ms	129 - 139 MB	CPU
BEVDet	TFLITE	float	Qualcomm® QCS8550 (Proxy)	1927.239 ms	103 - 105 MB	CPU
BEVDet	TFLITE	float	Qualcomm® SA8775P	2522.77 ms	128 - 139 MB	CPU
BEVDet	TFLITE	float	Qualcomm® QCS9075	2425.619 ms	126 - 1473 MB	CPU
BEVDet	TFLITE	float	Qualcomm® QCS8450 (Proxy)	2671.358 ms	129 - 149 MB	CPU
BEVDet	TFLITE	float	Qualcomm® SA7255P	3149.836 ms	129 - 139 MB	CPU
BEVDet	TFLITE	float	Qualcomm® SA8295P	2008.213 ms	87 - 95 MB	CPU
BEVDet	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1256.893 ms	75 - 85 MB	CPU
BEVDet	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1069.57 ms	89 - 100 MB	CPU

License

• The license for the original implementation of BEVDet can be found [here](https://github.com/HuangJunJie2017/BEVDet/blob/dev3.0/LICENSE https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf).

References

• BEVDet: High-Performance Multi-Camera 3D Object Detection in Bird-Eye-View
• 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.