Context Aggregation Network

This is the official Pytorch implementation code of paper Context aggregation network for semantic labeling in aerial images.

Requirements

Please refer to this file requirements.txt.

Getting started

1. Download ISPRS Vaihingen and Potsdam datasets on the website by following its instructions.
2. Put these datasets in corresponding dataset subfolder. Note that original colorful labels need to be converted to index-based (0,1,2,3,4,5) image using this code.
3. Run this command to train CAN model on ISPRS Vaihingen dataset, or Potsdam dataset by replacing isprs_vaihingen.yml with isprs_potsdam.yml. You can set many customized parameters in the .yml file.:

python train.py --config configs/isprs_vaihingen.yml

Models

Trained models used in benchmark evaluation for testing are provided in Mega.

Dataset folder

If you use my_loader.py as dataloader function, the dataset folder should have the following structure:

├── "dataset_name"                   
|   ├── train
|   ├── train_labels
|   ├── val
|   ├── val_labels
|   ├── test
|   ├── test_labels

Dependency

This repo is heavily based on the framework provided by pytorch-semseg. You can refer to that repo for more details.

Citation

If this is helpful for you, please consider to cite this article:

@article{cheng2019context,
  title={Context Aggregation Network for Semantic Labeling in Aerial Images},
  author={Cheng, Wensheng and Yang, Wen and Wang, Min and Wang, Gang and Chen, Jinyong},
  journal={Remote Sensing},
  volume={11},
  number={10},
  pages={1158},
  year={2019},
  publisher={Multidisciplinary Digital Publishing Institute}
}