VPSAM: Taming Segment Anything Model for Video Polyp Segmentation via Disentanglement and Spatio-temporal Side Network

Paper

VPSAM: Taming Segment Anything Model for Video Polyp Segmentation via Disentanglement and Spatio-temporal Side Network
Zhixue Fang, Yuzhi Liu, Huisi Wu, and Jing Qin
ECCV 2024

Due to the confidentiality agreement in commercial cooperation, we only provide codes of core modules and the whole trainable models for the convenience of comparisons.

Installation

conda create -n VPSAM python=3.8
conda activate VPSAM
pip install -r requirements.txt

Reference

@inproceedings{fang2024vpsam,
  title={VPSAM: Taming Segment Anything Model for Video Polyp Segmentation via Disentanglement and Spatio-temporal Side Network},
  author={Zhixue Fang, Yuzhi Liu, Huisi Wu, and Jing Qin},
  booktitle={Proceedings of the ECCV Conference on European Conference on Computer Vision},
  year={2024}
}