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 2.37, ONNX Runtime 1.23.0	Download
ONNX	w8a16	Universal	QAIRT 2.37, ONNX Runtime 1.23.0	Download
QNN_DLC	float	Universal	QAIRT 2.42	Download
QNN_DLC	w8a16	Universal	QAIRT 2.42	Download
TFLITE	float	Universal	QAIRT 2.42, TFLite 2.17.0	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® X Elite	4.702 ms	12 - 12 MB	NPU
Mobile-VIT	ONNX	float	Snapdragon® 8 Gen 3 Mobile	3.142 ms	0 - 168 MB	NPU
Mobile-VIT	ONNX	float	Qualcomm® QCS8550 (Proxy)	4.544 ms	0 - 133 MB	NPU
Mobile-VIT	ONNX	float	Qualcomm® QCS9075	5.619 ms	1 - 4 MB	NPU
Mobile-VIT	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	2.534 ms	0 - 139 MB	NPU
Mobile-VIT	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	2.145 ms	0 - 139 MB	NPU
Mobile-VIT	ONNX	w8a16	Snapdragon® X Elite	16.897 ms	16 - 16 MB	NPU
Mobile-VIT	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	15.32 ms	13 - 189 MB	NPU
Mobile-VIT	ONNX	w8a16	Qualcomm® QCS6490	339.344 ms	69 - 73 MB	CPU
Mobile-VIT	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	18.861 ms	7 - 17 MB	NPU
Mobile-VIT	ONNX	w8a16	Qualcomm® QCS9075	20.873 ms	12 - 14 MB	NPU
Mobile-VIT	ONNX	w8a16	Qualcomm® QCM6690	149.372 ms	63 - 74 MB	CPU
Mobile-VIT	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	12.17 ms	13 - 152 MB	NPU
Mobile-VIT	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	131.756 ms	63 - 73 MB	CPU
Mobile-VIT	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	11.754 ms	12 - 151 MB	NPU
Mobile-VIT	QNN_DLC	float	Snapdragon® X Elite	3.849 ms	1 - 1 MB	NPU
Mobile-VIT	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.456 ms	0 - 98 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	9.874 ms	1 - 59 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	3.48 ms	0 - 13 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® SA8775P	4.259 ms	1 - 61 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® QCS9075	4.47 ms	1 - 3 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	5.699 ms	0 - 94 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® SA7255P	9.874 ms	1 - 59 MB	NPU
Mobile-VIT	QNN_DLC	float	Qualcomm® SA8295P	6.44 ms	1 - 63 MB	NPU
Mobile-VIT	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.907 ms	1 - 68 MB	NPU
Mobile-VIT	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.611 ms	1 - 73 MB	NPU
Mobile-VIT	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.58 ms	0 - 108 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® QCS8275 (Proxy)	10.21 ms	0 - 85 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® QCS8550 (Proxy)	3.652 ms	0 - 19 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® SA8775P	4.461 ms	0 - 80 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® QCS9075	4.555 ms	0 - 15 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® QCS8450 (Proxy)	6.017 ms	0 - 94 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® SA7255P	10.21 ms	0 - 85 MB	NPU
Mobile-VIT	TFLITE	float	Qualcomm® SA8295P	6.691 ms	0 - 71 MB	NPU
Mobile-VIT	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	2.016 ms	0 - 74 MB	NPU
Mobile-VIT	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.644 ms	0 - 84 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.