Source-free Domain Adaptation Requires Penalized Diversity
This work addresses domain adaptation for scenarios where source data is unavailable, enhancing data privacy, but it is incremental as it builds on existing SFDA methods by refining diversity and regularization techniques.
The paper tackles the problem of source-free domain adaptation (SFDA) by proposing a novel algorithm called Penalized Diversity (PD) that uses separate feature extractors with distinct backbone architectures to increase representational diversity and introduces a Weak Hypothesis Penalization regularizer to mitigate weak hypothesis amplification, achieving improved adaptability across natural, synthetic, and medical domains.
While neural networks are capable of achieving human-like performance in many tasks such as image classification, the impressive performance of each model is limited to its own dataset. Source-free domain adaptation (SFDA) was introduced to address knowledge transfer between different domains in the absence of source data, thus, increasing data privacy. Diversity in representation space can be vital to a model`s adaptability in varied and difficult domains. In unsupervised SFDA, the diversity is limited to learning a single hypothesis on the source or learning multiple hypotheses with a shared feature extractor. Motivated by the improved predictive performance of ensembles, we propose a novel unsupervised SFDA algorithm that promotes representational diversity through the use of separate feature extractors with Distinct Backbone Architectures (DBA). Although diversity in feature space is increased, the unconstrained mutual information (MI) maximization may potentially introduce amplification of weak hypotheses. Thus we introduce the Weak Hypothesis Penalization (WHP) regularizer as a mitigation strategy. Our work proposes Penalized Diversity (PD) where the synergy of DBA and WHP is applied to unsupervised source-free domain adaptation for covariate shift. In addition, PD is augmented with a weighted MI maximization objective for label distribution shift. Empirical results on natural, synthetic, and medical domains demonstrate the effectiveness of PD under different distributional shifts.