---
library_name: pytorch
license: other
tags:
- android
pipeline_tag: video-object-tracking

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/track_anything/web-assets/model_demo.png)

# Track-Anything: Optimized for Qualcomm Devices

Track-Anything is a video based machine learning model to track an object in a video.

This is based on the implementation of Track-Anything found [here](https://github.com/gaomingqi/Track-Anything).
This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/track_anything) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).

Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) 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](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/track_anything/releases/v0.51.0/track_anything-onnx-float.zip)
| TFLITE | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/track_anything/releases/v0.51.0/track_anything-tflite-float.zip)

For more device-specific assets and performance metrics, visit **[Track-Anything on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/track_anything)**.


### Option 2: Export with Custom Configurations

Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/track_anything) 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 [Track-Anything on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/track_anything) for usage instructions.

## Model Details

**Model Type:** Model_use_case.video_object_tracking

**Model Stats:**
- Model checkpoint: xmem
- Input resolution: 320x568
- Number of parameters (encode_key_with_shrinkage): 9.72M
- Model size (encode_key_with_shrinkage) (float): 37.1 MB
- Number of parameters (encode_value): 23.3M
- Model size (encode_value) (float): 88.8 MB
- Number of parameters (encode_key_without_shrinkage): 9.71M
- Model size (encode_key_without_shrinkage) (float): 37.1 MB
- Number of parameters (segment): 30.1M
- Model size (segment) (float): 115 MB

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| encode_key_with_shrinkage | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.371 ms | 0 - 176 MB | NPU
| encode_key_with_shrinkage | ONNX | float | Snapdragon® X2 Elite | 2.809 ms | 18 - 18 MB | NPU
| encode_key_with_shrinkage | ONNX | float | Snapdragon® X Elite | 5.655 ms | 19 - 19 MB | NPU
| encode_key_with_shrinkage | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 3.677 ms | 0 - 202 MB | NPU
| encode_key_with_shrinkage | ONNX | float | Qualcomm® QCS8550 (Proxy) | 5.148 ms | 0 - 20 MB | NPU
| encode_key_with_shrinkage | ONNX | float | Qualcomm® QCS9075 | 9.443 ms | 14 - 17 MB | NPU
| encode_key_with_shrinkage | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 2.942 ms | 0 - 171 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 2.425 ms | 0 - 188 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 4.059 ms | 3 - 216 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 31.26 ms | 3 - 180 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 5.951 ms | 3 - 6 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® SA8775P | 9.434 ms | 3 - 181 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® QCS9075 | 10.706 ms | 2 - 29 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 11.696 ms | 3 - 199 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® SA7255P | 31.26 ms | 3 - 180 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Qualcomm® SA8295P | 10.298 ms | 3 - 163 MB | NPU
| encode_key_with_shrinkage | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 3.29 ms | 0 - 183 MB | NPU
| encode_key_without_shrinkage | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.494 ms | 0 - 215 MB | NPU
| encode_key_without_shrinkage | ONNX | float | Snapdragon® X2 Elite | 3.98 ms | 65 - 65 MB | NPU
| encode_key_without_shrinkage | ONNX | float | Snapdragon® X Elite | 6.953 ms | 66 - 66 MB | NPU
| encode_key_without_shrinkage | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 5.069 ms | 64 - 276 MB | NPU
| encode_key_without_shrinkage | ONNX | float | Qualcomm® QCS8550 (Proxy) | 6.848 ms | 58 - 77 MB | NPU
| encode_key_without_shrinkage | ONNX | float | Qualcomm® QCS9075 | 11.454 ms | 61 - 64 MB | NPU
| encode_key_without_shrinkage | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.364 ms | 63 - 262 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.542 ms | 0 - 223 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 5.579 ms | 19 - 234 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 33.515 ms | 20 - 203 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 7.706 ms | 20 - 23 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® SA8775P | 11.704 ms | 20 - 206 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® QCS9075 | 12.896 ms | 20 - 63 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 14.19 ms | 20 - 219 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® SA7255P | 33.515 ms | 20 - 203 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Qualcomm® SA8295P | 12.685 ms | 20 - 185 MB | NPU
| encode_key_without_shrinkage | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.386 ms | 19 - 208 MB | NPU
| encode_value | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 6.607 ms | 12 - 178 MB | NPU
| encode_value | ONNX | float | Snapdragon® X2 Elite | 6.048 ms | 39 - 39 MB | NPU
| encode_value | ONNX | float | Snapdragon® X Elite | 12.554 ms | 38 - 38 MB | NPU
| encode_value | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 9.886 ms | 1 - 206 MB | NPU
| encode_value | ONNX | float | Qualcomm® QCS8550 (Proxy) | 15.61 ms | 0 - 45 MB | NPU
| encode_value | ONNX | float | Qualcomm® QCS9075 | 18.841 ms | 6 - 14 MB | NPU
| encode_value | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 8.386 ms | 8 - 181 MB | NPU
| encode_value | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.306 ms | 2 - 201 MB | NPU
| encode_value | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 8.233 ms | 3 - 289 MB | NPU
| encode_value | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 51.329 ms | 3 - 217 MB | NPU
| encode_value | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 11.884 ms | 3 - 5 MB | NPU
| encode_value | TFLITE | float | Qualcomm® SA8775P | 17.049 ms | 3 - 215 MB | NPU
| encode_value | TFLITE | float | Qualcomm® QCS9075 | 19.168 ms | 0 - 55 MB | NPU
| encode_value | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 21.326 ms | 3 - 289 MB | NPU
| encode_value | TFLITE | float | Qualcomm® SA7255P | 51.329 ms | 3 - 217 MB | NPU
| encode_value | TFLITE | float | Qualcomm® SA8295P | 18.871 ms | 3 - 221 MB | NPU
| encode_value | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 6.837 ms | 3 - 217 MB | NPU
| segment | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 10.465 ms | 25 - 244 MB | NPU
| segment | ONNX | float | Snapdragon® X2 Elite | 11.733 ms | 38 - 38 MB | NPU
| segment | ONNX | float | Snapdragon® X Elite | 23.51 ms | 37 - 37 MB | NPU
| segment | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 17.483 ms | 24 - 267 MB | NPU
| segment | ONNX | float | Qualcomm® QCS8550 (Proxy) | 23.202 ms | 0 - 51 MB | NPU
| segment | ONNX | float | Qualcomm® QCS9075 | 38.679 ms | 21 - 45 MB | NPU
| segment | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 13.587 ms | 5 - 222 MB | NPU
| segment | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 11.447 ms | 0 - 195 MB | NPU
| segment | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 17.253 ms | 0 - 292 MB | NPU
| segment | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 103.297 ms | 2 - 195 MB | NPU
| segment | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 23.354 ms | 2 - 5 MB | NPU
| segment | TFLITE | float | Qualcomm® SA8775P | 33.308 ms | 0 - 191 MB | NPU
| segment | TFLITE | float | Qualcomm® QCS9075 | 38.29 ms | 0 - 83 MB | NPU
| segment | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 42.152 ms | 2 - 293 MB | NPU
| segment | TFLITE | float | Qualcomm® SA7255P | 103.297 ms | 2 - 195 MB | NPU
| segment | TFLITE | float | Qualcomm® SA8295P | 37.679 ms | 2 - 201 MB | NPU
| segment | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 13.211 ms | 1 - 201 MB | NPU

## License
* The license for the original implementation of Track-Anything can be found
  [here](https://github.com/gaomingqi/Track-Anything/blob/master/LICENSE).

## References
* [Track Anything: Segment Anything Meets Videos](https://arxiv.org/abs/2304.11968)
* [Source Model Implementation](https://github.com/gaomingqi/Track-Anything)

## Community
* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
* For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).