Mobile-VIT: Optimized for Qualcomm Devices

MobileVit is a machine learning model that can classify images from the Imagenet dataset. It can also be used as a backbone in building more complex models for specific use cases.

This is based on the implementation of Mobile-VIT 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 1.24.3	Download
ONNX	w8a16	Universal	QAIRT 2.42, ONNX Runtime 1.24.3	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.19.1	Download

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

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Model checkpoint: Imagenet
• Input resolution: 224x224
• Number of parameters: 5.57M
• Model size (float): 21.4 MB
• Model size (w8a16): 6.56 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Mobile-VIT	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.753 ms	1 - 85 MB	NPU
Mobile-VIT	ONNX	float	Snapdragon® X2 Elite	1.989 ms	12 - 12 MB	NPU
Mobile-VIT	ONNX	float	Snapdragon® X Elite	4.121 ms	12 - 12 MB	NPU
Mobile-VIT	ONNX	float	Snapdragon® 8 Gen 3 Mobile	2.609 ms	0 - 101 MB	NPU
Mobile-VIT	ONNX	float	Qualcomm® QCS8550 (Proxy)	3.732 ms	0 - 22 MB	NPU
Mobile-VIT	ONNX	float	Qualcomm® QCS9075	4.631 ms	1 - 4 MB	NPU
Mobile-VIT	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	2.082 ms	0 - 68 MB	NPU
Mobile-VIT	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	1.894 ms	0 - 108 MB	NPU
Mobile-VIT	ONNX	w8a16	Snapdragon® X2 Elite	2.03 ms	6 - 6 MB	NPU
Mobile-VIT	ONNX	w8a16	Snapdragon® X Elite	5.008 ms	6 - 6 MB	NPU
Mobile-VIT	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	3.073 ms	0 - 123 MB	NPU
Mobile-VIT	ONNX	w8a16	Qualcomm® QCS6490	320.342 ms	65 - 69 MB	CPU
Mobile-VIT	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	4.581 ms	0 - 12 MB	NPU
Mobile-VIT	ONNX	w8a16	Qualcomm® QCS9075	4.963 ms	0 - 3 MB	NPU
Mobile-VIT	ONNX	w8a16	Qualcomm® QCM6690	140.916 ms	64 - 75 MB	CPU
Mobile-VIT	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	2.284 ms	0 - 77 MB	NPU
Mobile-VIT	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	121.306 ms	65 - 76 MB	CPU
Mobile-VIT	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.58 ms	1 - 68 MB	NPU
Mobile-VIT	QNN_DLC	float	Snapdragon® X2 Elite	2.129 ms	1 - 1 MB	NPU
Mobile-VIT	QNN_DLC	float	Snapdragon® X Elite	3.799 ms	1 - 1 MB	NPU
Mobile-VIT	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.406 ms	0 - 94 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	9.806 ms	1 - 56 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	3.426 ms	1 - 102 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® SA8775P	4.209 ms	1 - 67 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® QCS9075	4.151 ms	1 - 3 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	5.597 ms	0 - 92 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® SA7255P	9.806 ms	1 - 56 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® SA8295P	6.441 ms	0 - 59 MB	NPU
Mobile-VIT	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.903 ms	1 - 57 MB	NPU
Mobile-VIT	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.696 ms	0 - 81 MB	NPU
Mobile-VIT	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.543 ms	0 - 109 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® QCS8275 (Proxy)	10.204 ms	0 - 77 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® QCS8550 (Proxy)	3.619 ms	0 - 7 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® SA8775P	4.458 ms	0 - 72 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® QCS9075	4.405 ms	0 - 15 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® QCS8450 (Proxy)	5.871 ms	0 - 100 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® SA7255P	10.204 ms	0 - 77 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® SA8295P	6.699 ms	0 - 63 MB	NPU
Mobile-VIT	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	2.01 ms	0 - 73 MB	NPU

License

• The license for the original implementation of Mobile-VIT can be found here.

References

• MOBILEVIT: LIGHT-WEIGHT, GENERAL-PURPOSE, AND MOBILE-FRIENDLY VISION TRANSFORMER
• 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.