This repository provides the code implementation for our paper [NeurIPS 2023] SPA: A Graph Spectral Alignment Perspective for Domain Adaptation.
Unsupervised domain adaptation (UDA) is a pivotal form in machine learning to extend the in-domain model to the distinctive target domains where the data distributions differ. Most prior works focus on capturing the inter-domain transferability but largely overlook rich intra-domain structures, which empirically results in even worse discriminability. In this work, we introduce a novel graph SPectral Alignment (SPA) framework to tackle the tradeoff. The core of our method is briefly condensed as follows: (i)-by casting the DA problem to graph primitives, SPA composes a coarse graph alignment mechanism with a novel spectral regularizer towards aligning the domain graphs in eigenspaces; (ii)-we further develop a fine-grained message propagation module — upon a novel neighbor-aware self-training mechanism — in order for enhanced discriminability in the target domain. On standardized benchmarks, the extensive experiments of SPA demonstrate that its performance has surpassed the existing cutting-edge DA methods. Coupled with dense model analysis, we conclude that our approach indeed possesses superior efficacy, robustness, discriminability, and transferability.
If this code is useful for your research, please refer to this paper for more details:
@inproceedings{xiao2023spa,
title={SPA: A Graph Spectral Alignment Perspective for Domain Adaptation},
author={Xiao, Zhiqing and Wang, Haobo and Jin, Ying and Feng, Lei and Chen, Gang and Huang, Fei and Zhao, Junbo},
journal={NeurIPS},
year={2023}
}
-
install packages
python == 3.7.13
pytorch == 1.10.1
torchvision == 0.11.2
sklearn == 1.0.2
numpy == 1.12.6
argparse, random, PIL
-
download dataset
- Office31
- OfficeHome
- VisDA2017
- DomainNet
-
run the train file with 'train_[uda/ssda]_[dataset].sh', e.g.
sh train_uda_officehome.sh spa 0.2 1.0 DANNE 0.3 laplac1 gauss