Contrastive Adaptation Network

This repo contains a modified version of Contrastive Adaptation Network (CAN) for time series datasets. We replaced the image feature extractor (ResNet) with the time-series compatible feature extractor from CoDATS. This allows a comparison of CAN with our time series contrastive domain adaptation method CALDA.

Dependencies

Below we list which packages and versions we used, though likely the exact versions are not required:

• Python 3.7.4
• PyTorch 1.9.0
• PyYAML 5.3.1
• torchvision 0.10.0
• torchinfo 1.5.3
• easydict 1.9
• pickle5 0.0.11

Datasets

For time series datasets, see CALDA instructions. That repository contains the scripts to generate the pickle files used by this code. Note CALDA should be cloned in ../calda, i.e. in the parent directory of this repo.

Training

For the time series training:

time ./experiments/scripts/train.sh ./experiments/config/timeseries/CAN/timeseries_train_train2val_cfg.yaml 0 CAN timeseries_train2val

The experiment log file and the saved checkpoints will be stored at ./experiments/ckpt/${experiment_name}

Test

For the time series best-target evaluation (i.e. using a comparable model selection methodology as CALDA):

time ./experiments/scripts/test_best_target.sh ./experiments/config/timeseries/timeseries_test_val_cfg.yaml 0 True timeseries_train2val timeseries_test ./experiments/ckpt

Citing

Please cite their paper if you use their code in your research:

@article{kangcontrastive,
  title={Contrastive Adaptation Network for Single-and Multi-Source Domain Adaptation},
  author={Kang, Guoliang and Jiang, Lu and Wei, Yunchao and Yang, Yi and Hauptmann, Alexander G},
  journal={IEEE transactions on pattern analysis and machine intelligence},
  year={2020}
}

@inproceedings{kang2019contrastive,
  title={Contrastive Adaptation Network for Unsupervised Domain Adaptation},
  author={Kang, Guoliang and Jiang, Lu and Yang, Yi and Hauptmann, Alexander G},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  pages={4893--4902},
  year={2019}
}

Thanks to third party

The way of setting configurations is inspired by https://github.com/rbgirshick/py-faster-rcnn.