Rethinking the Role of Pre-Trained Networks in Source-Free Domain Adaptation

TL;DR

This paper proposes a novel approach to source-free domain adaptation by integrating pre-trained networks into the adaptation process, using co-learning to enhance target pseudo-label quality and improve performance across multiple benchmarks.

Contribution

It introduces a co-learning strategy that leverages pre-trained networks during target adaptation, addressing overfitting and enhancing generalization in source-free domain adaptation.

Findings

Improved adaptation performance on 4 benchmark datasets.

Effective integration of pre-trained networks with existing SFDA methods.

Enhanced performance with stronger pre-trained models in the co-learning process.

Abstract

Source-free domain adaptation (SFDA) aims to adapt a source model trained on a fully-labeled source domain to an unlabeled target domain. Large-data pre-trained networks are used to initialize source models during source training, and subsequently discarded. However, source training can cause the model to overfit to source data distribution and lose applicable target domain knowledge. We propose to integrate the pre-trained network into the target adaptation process as it has diversified features important for generalization and provides an alternate view of features and classification decisions different from the source model. We propose to distil useful target domain information through a co-learning strategy to improve target pseudolabel quality for finetuning the source model. Evaluation on 4 benchmark datasets show that our proposed strategy improves adaptation performance and can be successfully integrated with existing SFDA methods. Leveraging modern pre-trained networks that have stronger representation learning ability in the co-learning strategy further boosts performance.