scott-yjyang / Vivim

Vivim: a Video Vision Mamba for Medical Video Lesion Segmentation

Vivim

Vivim: a Video Vision Mamba for Medical Video Object Segmentation

[arXiv]

News

• 24-03-11. ❗❗Update on Code. Welcome to taste.😄
• 24-02-08. Update on Method and Experiments.
• 24-01-26. This project is still quickly updating 🌝. Check TODO list to see what will be released next.
• 24-01-25. The paper has been released on arXiv.

A Quick Overview

Environment Setup

Clone this repository and navigate to the root directory of the project.

git clone https://github.com/scott-yjyang/Vivim.git

cd Vivim

Install basic package

conda env create -f environment.yml

Install casual-conv1d

cd causal-conv1d

python setup.py install

Install mamba

cd mamba

python setup.py install

TODO LIST

• Release Model
• Release training scripts
• Release evaluation
• Experiments on other video object segmentation datasets.
• configuration

Thanks

Code is based on ge-xing/SegMamba, hustvl/Vim, bowang-lab/U-Mamba.

Cite

If you find it useful, please cite

@article{yang2024vivim,
  title={Vivim: a Video Vision Mamba for Medical Video Object Segmentation},
  author={Yang, Yijun and Xing, Zhaohu and Zhu, Lei},
  journal={arXiv preprint arXiv:2401.14168},
  year={2024}
}