Deep noise-tolerant hashing for remote sensing image retrieval
This paper is accepted for publication with Signal Processing: Image Communication.
Before training, you need to download the UCMerced dataset http://weegee.vision.ucmerced.edu/datasets/landuse.html, AID dataset from https://captain-whu.github.io/AID ,WHURS dataset from https://captain-whu.github.io/BED4RS.
Pretrained model vit_small_patch16_224.pth is required for loading before training.
After the dataset has been prepared, we could run the follow command to train.
python main.py --is-train --dataset UCMerced --lr 0.0001 --wd 0.0004 --save-dir ./result/ucmd/16 --vit-path ./pretrained/vit_small_patch16_224.pth
@article{yan2025deep,
title={Deep noise-tolerant hashing for remote sensing image retrieval},
author={Yan, Chunyu and Wang, Lei and Qin, Qibing and Dai, Jiangyan and Zhang, Wenfeng},
journal={Signal Processing: Image Communication},
pages={117431},
year={2025},
publisher={Elsevier} }