Cross Contrasting Feature Perturbation for Domain Generalization

TL;DR

This paper introduces a novel online feature perturbation framework for domain generalization that improves model robustness to unseen domains without relying on generative models or domain labels.

Contribution

It proposes a learnable, semantic-consistent feature perturbation method that simulates domain shifts in the latent space, outperforming previous state-of-the-art methods.

Findings

Outperforms previous methods on DomainBed benchmark

Effectively alleviates domain shift in OOD scenarios

Does not require generative models or domain labels

Abstract

Domain generalization (DG) aims to learn a robust model from source domains that generalize well on unseen target domains. Recent studies focus on generating novel domain samples or features to diversify distributions complementary to source domains. Yet, these approaches can hardly deal with the restriction that the samples synthesized from various domains can cause semantic distortion. In this paper, we propose an online one-stage Cross Contrasting Feature Perturbation (CCFP) framework to simulate domain shift by generating perturbed features in the latent space while regularizing the model prediction against domain shift. Different from the previous fixed synthesizing strategy, we design modules with learnable feature perturbations and semantic consistency constraints. In contrast to prior work, our method does not use any generative-based models or domain labels. We conduct extensive experiments on a standard DomainBed benchmark with a strict evaluation protocol for a fair comparison. Comprehensive experiments show that our method outperforms the previous state-of-the-art, and quantitative analyses illustrate that our approach can alleviate the domain shift problem in out-of-distribution (OOD) scenarios.