Towards Domain-Specific Features Disentanglement for Domain Generalization
This work addresses domain generalization for machine learning systems facing distributional shifts, presenting an incremental improvement by focusing on previously overlooked domain-specific features.
The paper tackles the problem of domain generalization by addressing irrelevant domain-specific features that hinder cross-domain learning, proposing a contrastive-based disentanglement method (CDDG) that outperforms state-of-the-art approaches on benchmark datasets.
Distributional shift between domains poses great challenges to modern machine learning algorithms. The domain generalization (DG) signifies a popular line targeting this issue, where these methods intend to uncover universal patterns across disparate distributions. Noted, the crucial challenge behind DG is the existence of irrelevant domain features, and most prior works overlook this information. Motivated by this, we propose a novel contrastive-based disentanglement method CDDG, to effectively utilize the disentangled features to exploit the over-looked domain-specific features, and thus facilitating the extraction of the desired cross-domain category features for DG tasks. Specifically, CDDG learns to decouple inherent mutually exclusive features by leveraging them in the latent space, thus making the learning discriminative. Extensive experiments conducted on various benchmark datasets demonstrate the superiority of our method compared to other state-of-the-art approaches. Furthermore, visualization evaluations confirm the potential of our method in achieving effective feature disentanglement.