Skip to content

Source code for Self-supervised Likelihood Estimation with Energy Guidance for Anomaly Segmentation in Urban Scenes (AAAI 2024)

Notifications You must be signed in to change notification settings

yuanpengtu/SLEEG

Repository files navigation

SLEEG

Official implementation of our AAAI paper: Self-supervised Likelihood Estimation with Energy Guidance for Anomaly Segmentation in Urban Scenes

Self-supervised Likelihood Estimation with Energy Guidance for Anomaly Segmentation in Urban Scenes (AAAI 2024)

This is the pytorch implementation of the paper (accepted by AAAI 2024).

Performance

Installation

For the training set and test set, please refer to link to download cityscapes and fishyscapes datasets. Place the downloaded dataset at ./data/.

The docker file for running our SLEEG can be found in ./docker, where the submission file of our SLEEG on the Fishyscapes dataset is also available.

Training

bash scripts/train.sh

Inference

bash scripts/static.sh

The trained weights of SLEEG are available at link.

Our results on the leaderboard of Fishyscapes can be found in link, where anonymous submission (mall) is our SLEEG.

Acknowledgement & Citation

The code is built on mmseg. Many thanks for their great work. If you find this repo useful for your research, please consider citing our paper:

@misc{tu2023selfsupervised,
      title={Self-supervised Likelihood Estimation with Energy Guidance for Anomaly Segmentation in Urban Scenes}, 
      author={Yuanpeng Tu and Yuxi Li and Boshen Zhang and Liang Liu and Jiangning Zhang and Yabiao Wang and Chengjie Wang and Cai Rong Zhao},
      year={2023},
      eprint={2302.06815},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

About

Source code for Self-supervised Likelihood Estimation with Energy Guidance for Anomaly Segmentation in Urban Scenes (AAAI 2024)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages