Mixstyle-Entropy: Domain Generalization with Causal Intervention and Perturbation

TL;DR

This paper introduces InPer, a causal intervention and perturbation framework that improves deep neural network generalization across unseen domains by refining causal variables during training and testing.

Contribution

The paper presents a novel holistic approach combining causal intervention and perturbation, enhancing domain generalization in neural networks.

Findings

InPer outperforms existing methods on multiple cross-domain tasks.

Entropy-based causal intervention improves causal variable selection.

Causal perturbation with homeostatic score enhances test-time robustness.

Abstract

Despite the considerable advancements achieved by deep neural networks, their performance tends to degenerate when the test environment diverges from the training ones. Domain generalization (DG) solves this issue by learning representations independent of domain-related information, thus facilitating extrapolation to unseen environments. Existing approaches typically focus on formulating tailored training objectives to extract shared features from the source data. However, the disjointed training and testing procedures may compromise robustness, particularly in the face of unforeseen variations during deployment. In this paper, we propose a novel and holistic framework based on causality, named InPer, designed to enhance model generalization by incorporating causal intervention during training and causal perturbation during testing. Specifically, during the training phase, we employ entropy-based causal intervention (EnIn) to refine the selection of causal variables. To identify samples with anti-interference causal variables from the target domain, we propose a novel metric, homeostatic score, through causal perturbation (HoPer) to construct a prototype classifier in test time. Experimental results across multiple cross-domain tasks confirm the efficacy of InPer.