A Source-Free Approach for Domain Adaptation via Multiview Image Transformation and Latent Space Consistency

TL;DR

This paper presents a novel source-free domain adaptation method that leverages multiview image transformation and latent space consistency to learn domain-invariant features directly from the target domain, avoiding source data and complex alignment.

Contribution

It introduces a new approach using multiview augmentation and latent space consistency for source-free domain adaptation, eliminating the need for source data or pseudo-label refinement.

Findings

Achieves high classification accuracy on multiple datasets.

Improves existing methods' accuracy by up to 7.26%.

Uses a ConvNeXt-based encoder with combined loss functions.

Abstract

Domain adaptation (DA) addresses the challenge of transferring knowledge from a source domain to a target domain where image data distributions may differ. Existing DA methods often require access to source domain data, adversarial training, or complex pseudo-labeling techniques, which are computationally expensive. To address these challenges, this paper introduces a novel source-free domain adaptation method. It is the first approach to use multiview augmentation and latent space consistency techniques to learn domain-invariant features directly from the target domain. Our method eliminates the need for source-target alignment or pseudo-label refinement by learning transferable representations solely from the target domain by enforcing consistency between multiple augmented views in the latent space. Additionally, the method ensures consistency in the learned features by generating multiple augmented views of target domain data and minimizing the distance between their feature representations in the latent space. We also introduce a ConvNeXt-based encoder and design a loss function that combines classification and consistency objectives to drive effective adaptation directly from the target domain. The proposed model achieves an average classification accuracy of 90. 72\%, 84\%, and 97. 12\% in Office-31, Office-Home and Office-Caltech datasets, respectively. Further evaluations confirm that our study improves existing methods by an average classification accuracy increment of +1.23\%, +7.26\%, and +1.77\% on the respective datasets.