CVT: Optimized for Qualcomm Devices

Cross-View Transformer generates real-time bird's-eye view maps from multiple vehicle cameras for autonomous driving.

This is based on the implementation of CVT 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 1.27.1	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 CVT 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 CVT on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.driver_assistance

Model Stats:

• Inference latency: RealTime
• Input resolution: 1x6x3x224x480
• Model checkpoint: vehicles_50k.pt
• Model size (float): 5.18 MB
• Number of parameters: 1.33M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
CVT	ONNX	float	Snapdragon® X2 Elite	182.309 ms	8 - 8 MB	NPU
CVT	ONNX	float	Snapdragon® X Elite	277.135 ms	20 - 20 MB	NPU
CVT	ONNX	float	Snapdragon® 8 Gen 3 Mobile	228.014 ms	8 - 2726 MB	NPU
CVT	ONNX	float	Snapdragon® 8 Gen 1 Mobile	441.492 ms	8 - 2755 MB	NPU
CVT	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	330.321 ms	8 - 12 MB	NPU
CVT	ONNX	float	Qualcomm® QCS8450	441.492 ms	8 - 2755 MB	NPU
CVT	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	277.135 ms	20 - 20 MB	NPU
CVT	ONNX	float	Qualcomm® Dragonwing™ Q-8750	189.326 ms	1 - 2396 MB	NPU
CVT	ONNX	float	Snapdragon® 8 Elite Mobile	189.326 ms	1 - 2396 MB	NPU
CVT	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	167.613 ms	3 - 2444 MB	NPU
CVT	QNN_DLC	float	Snapdragon® X2 Elite	183.488 ms	7 - 7 MB	NPU
CVT	QNN_DLC	float	Snapdragon® X Elite	282.582 ms	7 - 7 MB	NPU
CVT	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	230.648 ms	7 - 2575 MB	NPU
CVT	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	435.352 ms	7 - 2720 MB	NPU
CVT	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	257.47 ms	7 - 17 MB	NPU
CVT	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	337.434 ms	8 - 12 MB	NPU
CVT	QNN_DLC	float	Qualcomm® SA8775P	292.089 ms	2 - 2213 MB	NPU
CVT	QNN_DLC	float	Qualcomm® SA8650P	292.089 ms	2 - 2213 MB	NPU
CVT	QNN_DLC	float	Qualcomm® SA8255P	292.089 ms	2 - 2213 MB	NPU
CVT	QNN_DLC	float	Qualcomm® QCS8450	435.352 ms	7 - 2720 MB	NPU
CVT	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	282.824 ms	7 - 17 MB	NPU
CVT	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	282.582 ms	7 - 7 MB	NPU
CVT	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	184.9 ms	0 - 2289 MB	NPU
CVT	QNN_DLC	float	Qualcomm® SA7255P	424.554 ms	1 - 2212 MB	NPU
CVT	QNN_DLC	float	Qualcomm® SA8295P	333.7 ms	0 - 2324 MB	NPU
CVT	QNN_DLC	float	Snapdragon® 8 Elite Mobile	184.9 ms	0 - 2289 MB	NPU
CVT	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	167.064 ms	7 - 2319 MB	NPU
CVT	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	218.548 ms	0 - 2673 MB	NPU
CVT	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	393.463 ms	1 - 2727 MB	NPU
CVT	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	325.239 ms	0 - 473 MB	NPU
CVT	TFLITE	float	Qualcomm® QCS8450	393.463 ms	1 - 2727 MB	NPU
CVT	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	182.78 ms	1 - 2363 MB	NPU
CVT	TFLITE	float	Qualcomm® SA8295P	330.46 ms	1 - 2306 MB	NPU
CVT	TFLITE	float	Snapdragon® 8 Elite Mobile	182.78 ms	1 - 2363 MB	NPU
CVT	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	165.886 ms	0 - 2314 MB	NPU

License

• The license for the original implementation of CVT can be found here.

References

• Cross-view Transformers for real-time Map-view Semantic Segmentation
• 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.