The repo of the ECCV 2020 paper: Progressively Guided Alternate Refinement Network for RGB-D Salient Object Detection.
Depth format: nearer pixels appear brighter and vice versa. Download our pre-trained model Baidu(9rr7) and Google and put it into $models/
. Then, run
python3 test.py
We provide saliency maps of our PGAR on 7 datasets: Baidu(3jzx) and Google.
@inproceedings{chen2020eccv,
author={Chen, Shuhan and Fu, Yun},
booktitle={European Conference on Computer Vision},
title={Progressively Guided Alternate Refinement Network for RGB-D Salient Object Detection},
year={2020}
}