Awesome
PEBAL
<img width="933" alt="Screen Shot 2022-06-11 at 2 56 11 pm" src="https://user-images.githubusercontent.com/102338056/173174121-f515ce6d-a865-4dcd-aa00-4b0e4fd5d448.png"> <!-- ![image](https://user-images.githubusercontent.com/19222962/161691512-61a2dfa8-2079-465c-abaa-5b8fdf42e5f7.png) -->[ECCV'22 Oral] Pixel-wise Energy-biased Abstention Learning for Anomaly Segmentation on Complex Urban Driving Scenes
by Yu Tian*, Yuyuan Liu*, Guansong Pang, Fengbei Liu, Yuanhong Chen, Gustavo Carneiro.
Update
- :sparkles: Results on Segment-Me-if-You-Can has been released!
- :beers: Our newest work RPL for anomaly segmentation has been accepted in ICCV'23!
Installation
Please install the dependencies and dataset based on this installation document.
Getting started
Please follow this instruction document to reproduce our results.
Acknowledgement & Citation
The code is partially borrowed from CPS. Many thanks for their great work.
If you find this repo useful for your research, please consider citing our paper:
@misc{tian2021pixelwise,
title={Pixel-wise Energy-biased Abstention Learning for Anomaly Segmentation on Complex Urban Driving Scenes},
author={Yu Tian and Yuyuan Liu and Guansong Pang and Fengbei Liu and Yuanhong Chen and Gustavo Carneiro},
year={2021},
eprint={2111.12264},
archivePrefix={arXiv},
primaryClass={cs.CV}
}