Documentation: https://mmtracking.readthedocs.io/

Introduction

MMTracking is an open source video perception toolbox based on PyTorch. It is a part of the OpenMMLab project.

The master branch works with PyTorch 1.3 to 1.7.

Major features

• The First Unified Video Perception Platform

  We are the first open source toolbox that unifies versatile video perception tasks include video object detection, single object tracking, and multiple object tracking.

• Modular Design

  We decompose the video perception framework into different components and one can easily construct a customized method by combining different modules.

• Simple, Fast and Strong

  Simple: MMTracking interacts with other OpenMMLab projects. It is built upon MMDetection that we can capitalize any detector only through modifying the configs.

  Fast: All operations run on GPUs. The training and inference speeds are faster than or comparable to other implementations.

  Strong: We reproduce state-of-the-art models and some of them even outperform the offical implementations.

License

This project is released under the Apache 2.0 license.

Changelog

v0.5.0 was released in 04/01/2021. Please refer to changelog.md for details and release history.

Benchmark and model zoo

Results and models are available in the model zoo.

Supported methods of video object detection:

• DFF
• FGFA
• SELSA

Supported methods of multi object tracking:

• SORT/DeepSORT
• Tracktor

Supported methods of single object tracking:

• SiameseRPN++

Installation

Please refer to install.md for install instructions.

Get Started

Please see dataset.md and quick_run.md for the basic usage of MMTracking. We also provide usage tutorials.

Contributing

We appreciate all contributions to improve MMTracking. Please refer to CONTRIBUTING.md for the contributing guideline.

Acknowledgement

MMTracking is an open source project that welcome any contribution and feedback. We wish that the toolbox and benchmark could serve the growing research community by providing a flexible as well as standardized toolkit to reimplement existing methods and develop their own new video perception methods.