Python 3.7.0
Pytorch 1.6.0
Visdom 0.1.8.9
Torchvision 0.7.0
- CD2014 dataset
- paper: changedetection.net: A new change detection benchmark dataset
- paper: CDnet 2014: An Expanded Change Detection Benchmark Dataset
- dataset: http://changedetection.net/
- if the above link didn't work, please try this baidudisk link: [Baiduyun] the password is: ob6j
- VL-CMU-CD
- paper: Street-view change detection with deconvolutional networks
- dataset: https://ghsi.github.io/proj/RSS2016.html
- if the above link didn't work, please try this baidudisk link: [Baiduyun] the password is: v82c
Pretrained models for CDNet-2014 and VL-CMU-CD are available. You can download them from the following link.
- CDNet-2014: [Baiduyun] the password is 78cp. [GoogleDrive].
- We uploaded six models trained on CDNet-2014 dataset, they are SViT_E1_D1_16, SViT_E1_D1_32, SViT_E4_D4_16, SViT_E4_D4_32, Res_SViT_E1_D1_16 and Res_SViT_E4_D4_16.
- VL-CMU-CD: [Baiduyun] the password is ydzl. [GoogleDrive].
- We uploaded four models trained on VL-CMU-CD dataset, they are SViT_E1_D1_16, SViT_E1_D1_32, Res_SViT_E1_D1_16 and Res_SViT_E1_D1_32.
Before test, please download datasets and pretrained models. Copy pretrained models to folder './dataset_name/outputs/best_weights', and run the following command:
cd TransCD_ROOT
python test.py --net_cfg <net name> --train_cfg <training configuration>
Use --save_changemap True to save predicted changemaps.
For example:
python test.py --net_cfg SViT_E1_D1_32 --train_cfg CDNet_2014 --save_changemap True
Before training, please download datasets and revise dataset path in configs.py to your path. CD TransCD_ROOT
python -m visdom.server
python train.py --net_cfg <net name> --train_cfg <training configuration>
For example:
python -m visdom.server
python train.py --net_cfg Res_SViT_E1_D1_16 --train_cfg VL_CMU_CD
To display training processing, open 'http://localhost:8097' in your browser.
If you use this repository or would like to refer the paper, please use the following BibTex entry.
@article{Wang:21,
author = {Zhixue Wang and Yu Zhang and Lin Luo and Nan Wang},
journal = {Opt. Express},
keywords = {Feature extraction; Neural networks; Object detection; Segmentation; Spatial resolution; Vision modeling},
number = {25},
pages = {41409--41427},
publisher = {OSA},
title = {TransCD: scene change detection via transformer-based architecture},
volume = {29},
month = {Dec},
year = {2021},
url = {http://www.osapublishing.org/oe/abstract.cfm?URI=oe-29-25-41409},
doi = {10.1364/OE.440720},
}
-Akcay, Samet, Amir Atapour-Abarghouei, and Toby P. Breckon. "Ganomaly: Semi-supervised anomaly detection via adversarial training." Asian conference on computer vision. Springer, Cham, 2018.
-Chen, Jieneng, et al. "Transunet: Transformers make strong encoders for medical image segmentation." arXiv preprint arXiv:2102.04306 (2021).