CorDA

Code for our paper Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimation

Prerequisite

Please create and activate the following conda envrionment.

# It may take several minutes for conda to solve the environment
conda update conda
conda env create -f environment.yml
conda activate corda

Code was tested on a V100 with 16G Memory. To reproduce our results, please kindly create and use this environment.

Train a CorDA model

# Train for the SYNTHIA2Cityscapes task
bash run_synthia_stereo.sh
# Train for the GTA2Cityscapes task
bash run_gta.sh

Test the trained model

bash shells/eval_syn2city.sh
bash shells/eval_gta2city.sh

Pre-trained models are provided (Google Drive). Please put them in ./checkpoint.

  • The provided SYNTHIA2Cityscapes model achieves 56.3 mIoU (16 classes) at the end of the training.
  • The provided GTA2Cityscapes model achieves 57.7 mIoU (19 classes) at the end of the training.

Reported Results on SYNTHIA2Cityscapes

Method mIoU*(13) mIoU(16)
CBST 48.9 42.6
FDA 52.5 -
DADA 49.8 42.6
DACS 54.8 48.3
CorDA 62.8 55.0

Citation

Please cite our work if you find it useful.

@article{wang2021domain,
  title={Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimation},
  author={Wang, Qin and Dai, Dengxin and Hoyer, Lukas and Fink, Olga and Van Gool, Luc},
  journal={arXiv preprint arXiv:2104.13613},
  year={2021}
}

Acknowledgement

  • DACS is used as our codebase and our DA baseline official
  • SFSU as the source of stereo Cityscapes depth estimation Official

For questions regarding the code, please contact [email protected] .

GitHub

https://github.com/qinenergy/corda