Code release for "Domain Generalization via Frequency-domain-based Feature Disentanglement and Interaction" (ACM MM 2022): https://arxiv.org/abs/2201.08029.
Part of the code is inherited from Episodic-DG.
GPU GeForce RTX 1080 Ti
pytorch==1.9.0
torchvision==0.10.0
cudatoolkit==10.2.89
Please download the PACS datasets and use the official train/val split.
We use the pytorch pretrained ResNet-18 model from https://download.pytorch.org/models/resnet18-5c106cde.pth.
- Train from scratch with command:
CUDA_VISIBLE_DEVICES=0,1,2,3 python main_agg.py
Please cite the related works in your publications if it helps your research:
@inproceedings{wang2022domain,
title={Domain Generalization via Frequency-domain-based Feature Disentanglement and Interaction},
author={Wang, Jingye and Du, Ruoyi and Chang, Dongliang and Liang, KongMing and Ma, Zhanyu},
booktitle={Proceedings of the 30th ACM International Conference on Multimedia},
year={2022}
}