Skip to content

Latest commit

 

History

History
38 lines (30 loc) · 1.24 KB

README.md

File metadata and controls

38 lines (30 loc) · 1.24 KB
Raw

ViCatDA

Code release for Vicinal and categorical domain adaptation, which is published by Pattern Recognition in 2021.

Project Page $\cdot$ PDF Download

The paper is available here or at the arXiv archive.

Requirements

  • python 3.6.4
  • pytorch 1.4.0
  • torchvision 0.5.0

Data preparation

The structure of the used datasets is shown in the folder ./data/datasets/.

The original datasets can be downloaded here.

Model training

  1. Replace paths and domains in run.sh with those in one's own system.
  2. Install necessary python packages.
  3. Run command sh run.sh.

The results are saved in the folder ./checkpoints/.

Article citation

@article{vicatda,
author = {Hui Tang and Kui Jia},
title = {Vicinal and categorical domain adaptation},
journal = {Pattern Recognition},
year = {2021},
volume = {115},
pages = {107907},
issn = {0031-3203},
doi = {https://doi.org/10.1016/j.patcog.2021.107907},
url = {https://www.sciencedirect.com/science/article/pii/S0031320321000947},
}