Semi-Auto-Multi-Level-Annotation-Tool

This repo is the office implement of the Semi-Auto Multi-Level Annotation Tool used in M-cube-VOS. It allows for user to annotate the mask of target objects efficiently. In M-cube-VOS, The pipeline of data collection is as follow:

News 🔥

• We release the Annotation Tool.

• M-cube-VOS get accepted in CVPR 2025.

• We release the dataset M-cube-VOS in baidu disk.

Installation

Our test environment is :

• Ubuntu 20.04.6 LTS

• Python 3.8.19

• torch 2.3.1+cu118 , torchaudio 2.3.1+cu118, torchvision 0.18.1+cu118

tip: The machine running this tool is expected to need GeForce GTX and RTX.

Clone our repository:

git clone https://github.com/Lijiaxin0111/SemiAuto-Multi-Level-Annotation-Tool.git   

Create Environment:

conda create -n SemiAuto_AnnotateTool python=3.8
conda activate SemiAuto_AnnotateTool

Install with pip:

cd SemiAuto-Multi-Level-Annotation-Tool

pip install torch==2.3.1 torchvision==0.18.1 torchaudio==2.3.1 --index-url https://download.pytorch.org/whl/cu118

pip install -r requirements.txt

(If you encounter the File "setup.py" not found error, upgrade your pip with pip install --upgrade pip)

(If you encounter "error: Microsoft Visual C++ 14.0 or greater is required.", get it with "Microsoft C++ Build Tools": https://visualstudio.microsoft.com/visual-cpp-build-tools/)

Usage

Quick Start

Tips: If you are running this on a remote server, you can use VNC or X11 forwarding.

python interactive_demo.py  --video ./demo_data/make_glass.mp4  --workspace ./workspace/make_glass --num_objects 1 --gpu 0

More Usage

• 中文文档
• English Doc

UI dmeo

Citation

@InProceedings{chen2024m3vos_2025_CVPR,
    author    = {Zixuan Chen and Jiaxin Li and Liming Tan and Yejie Guo and Junxuan Liang and Cewu Lu and Yong-Lu Li},
    title     = {M$^3$-VOS: Multi-Phase, Multi-Transition, and Multi-Scenery Video Object Segmentation},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2025}
}

References

• This Semi-Auto-Multi-Level-Annotation-Tool is based on Cutie GUI tool, IVS, MiVOS, and XMem.

  • The Cutie GUI tools uses RITM for interactive image segmentation. This repository also contains a redistribution of their code in gui/ritm. That part of code follows RITM's license.

  • For automatic video segmentation/integration with external detectors, see DEVA.

  • Cutie GUI tool used ProPainter in the video inpainting demo.

• Thanks to Cutie, RTIM, XMem++, IVS, MiVOS, and XMem for making this possible.

License

This project is licensed under the MIT License . You are free to use, modify, and distribute the code, provided that the original copyright notice and license are included.