SF(DA)$^2$: Source-free Domain Adaptation Through the Lens of Data Augmentation

TL;DR

This paper introduces SF(DA)$^2$, a novel source-free domain adaptation method that uses data augmentation in feature space with graph-based clustering and regularization to improve adaptation without high computational costs.

Contribution

It proposes a new data augmentation approach in feature space using graph construction, spectral clustering, and regularizers, addressing challenges in source-free domain adaptation.

Findings

Superior performance on 2D image datasets

Effective adaptation on 3D point cloud data

Handles highly imbalanced datasets

Abstract

In the face of the deep learning model's vulnerability to domain shift, source-free domain adaptation (SFDA) methods have been proposed to adapt models to new, unseen target domains without requiring access to source domain data. Although the potential benefits of applying data augmentation to SFDA are attractive, several challenges arise such as the dependence on prior knowledge of class-preserving transformations and the increase in memory and computational requirements. In this paper, we propose Source-free Domain Adaptation Through the Lens of Data Augmentation (SF(DA)$^2$), a novel approach that leverages the benefits of data augmentation without suffering from these challenges. We construct an augmentation graph in the feature space of the pretrained model using the neighbor relationships between target features and propose spectral neighborhood clustering to identify partitions in the prediction space. Furthermore, we propose implicit feature augmentation and feature disentanglement as regularization loss functions that effectively utilize class semantic information within the feature space. These regularizers simulate the inclusion of an unlimited number of augmented target features into the augmentation graph while minimizing computational and memory demands. Our method shows superior adaptation performance in SFDA scenarios, including 2D image and 3D point cloud datasets and a highly imbalanced dataset.