Polycentric Clustering and Structural Regularization for Source-free Unsupervised Domain Adaptation
Xinyu Guan (Nanjing University of Aeronautics and Astronautics), Han Sun (NUAA),* Ningzhong Liu (Nanjing University of Aeronautics and Astronautics), Huiyu Zhou (University of Leicester)
The 33rd British Machine Vision Conference
Abstract
Source-Free Domain Adaptation (SFDA) aims to solve the domain adaptation problem by transferring the knowledge learned from a pre-trained source model to an unseen target domain. Most existing methods assign pseudo-labels to target data by generating feature prototypes. However, due to the discrepancy in data distribution between the source domain and the target domain and the category imbalance in the target domain, there are severe class-biased in the generated feature prototypes and noisy pseudo-labels. Besides, the data structure of the target domain is often ignored, which is crucial for clustering. In this paper, a novel framework named PCSR is proposed to tackle SFDA via intra-class Polycentric Clustering and Structural Regularization strategy. Firstly, an inter-class balanced sampling strategy is proposed to generate representative feature prototypes for each class. Furthermore, k-means clustering is introduced to generate multiple clustering centers for each class in the target domain to obtain robust pseudo labels. Finally, to enhance the model's generalization, structural regularization is designed for the target domain. Extensive experiments on three UDA benchmark datasets show that our method outperforms and is competitive to the state-of-the-art, demonstrating our approach's superiority for visual domain adaptation problems.Video
Citation
@inproceedings{Guan_2022_BMVC,
author = {Xinyu Guan and Han Sun and Ningzhong Liu and Huiyu Zhou},
title = {Polycentric Clustering and Structural Regularization for Source-free Unsupervised Domain Adaptation},
booktitle = {33rd British Machine Vision Conference 2022, {BMVC} 2022, London, UK, November 21-24, 2022},
publisher = {{BMVA} Press},
year = {2022},
url = {https://bmvc2022.mpi-inf.mpg.de/0485.pdf}
}
Copyright © 2022
The British Machine Vision Conference is organised by