MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer
Paper • 2110.02178 • Published • 3
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 1.24.1 | Download |
| ONNX | w8a16 | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | Download |
| QNN_DLC | float | Universal | QAIRT 2.43 | Download |
| QNN_DLC | w8a16 | Universal | QAIRT 2.43 | Download |
| TFLITE | float | Universal | QAIRT 2.43, 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® X2 Elite | 1.89 ms | 12 - 12 MB | NPU |
| Mobile-VIT | ONNX | float | Snapdragon® X Elite | 3.956 ms | 12 - 12 MB | NPU |
| Mobile-VIT | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 2.523 ms | 0 - 110 MB | NPU |
| Mobile-VIT | ONNX | float | Qualcomm® QCS8550 (Proxy) | 3.597 ms | 0 - 16 MB | NPU |
| Mobile-VIT | ONNX | float | Qualcomm® QCS9075 | 4.693 ms | 1 - 4 MB | NPU |
| Mobile-VIT | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 1.979 ms | 0 - 89 MB | NPU |
| Mobile-VIT | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 1.668 ms | 0 - 102 MB | NPU |
| Mobile-VIT | ONNX | w8a16 | Snapdragon® X2 Elite | 2.076 ms | 7 - 7 MB | NPU |
| Mobile-VIT | ONNX | w8a16 | Snapdragon® X Elite | 5.091 ms | 8 - 8 MB | NPU |
| Mobile-VIT | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 3.128 ms | 0 - 144 MB | NPU |
| Mobile-VIT | ONNX | w8a16 | Qualcomm® QCS6490 | 325.906 ms | 65 - 69 MB | CPU |
| Mobile-VIT | ONNX | w8a16 | Qualcomm® QCS8550 (Proxy) | 4.64 ms | 0 - 10 MB | NPU |
| Mobile-VIT | ONNX | w8a16 | Qualcomm® QCS9075 | 5.094 ms | 0 - 3 MB | NPU |
| Mobile-VIT | ONNX | w8a16 | Qualcomm® QCM6690 | 141.041 ms | 63 - 73 MB | CPU |
| Mobile-VIT | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 2.336 ms | 0 - 80 MB | NPU |
| Mobile-VIT | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 122.873 ms | 65 - 76 MB | CPU |
| Mobile-VIT | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 1.918 ms | 0 - 115 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Snapdragon® X2 Elite | 2.125 ms | 1 - 1 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Snapdragon® X Elite | 3.851 ms | 1 - 1 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 2.457 ms | 0 - 99 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 9.916 ms | 1 - 68 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 3.494 ms | 1 - 3 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Qualcomm® SA8775P | 4.271 ms | 1 - 70 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Qualcomm® QCS9075 | 4.485 ms | 3 - 5 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 5.641 ms | 0 - 94 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Qualcomm® SA7255P | 9.916 ms | 1 - 68 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Qualcomm® SA8295P | 6.47 ms | 1 - 62 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 1.925 ms | 0 - 73 MB | NPU |
| Mobile-VIT | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 1.603 ms | 1 - 72 MB | NPU |
| Mobile-VIT | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 2.586 ms | 0 - 108 MB | NPU |
| Mobile-VIT | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 10.216 ms | 0 - 81 MB | NPU |
| Mobile-VIT | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 3.672 ms | 0 - 13 MB | NPU |
| Mobile-VIT | TFLITE | float | Qualcomm® SA8775P | 4.517 ms | 0 - 83 MB | NPU |
| Mobile-VIT | TFLITE | float | Qualcomm® QCS9075 | 4.575 ms | 0 - 15 MB | NPU |
| Mobile-VIT | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 6.002 ms | 0 - 93 MB | NPU |
| Mobile-VIT | TFLITE | float | Qualcomm® SA7255P | 10.216 ms | 0 - 81 MB | NPU |
| Mobile-VIT | TFLITE | float | Qualcomm® SA8295P | 6.691 ms | 0 - 69 MB | NPU |
| Mobile-VIT | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.022 ms | 0 - 81 MB | NPU |
| Mobile-VIT | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 1.654 ms | 0 - 86 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.