Neuron Coverage-Guided Domain Generalization

TL;DR

This paper introduces a neuron coverage-guided approach to improve domain generalization in deep neural networks, especially when training data is limited to a single domain, by maximizing neuron coverage and using gradient similarity regularization.

Contribution

It proposes a novel neuron coverage-based training method with gradient regularization to enhance DNN generalization to out-of-distribution samples in single-domain scenarios.

Findings

Outperforms state-of-the-art methods on various domain generalization tasks.

Neuron coverage maximization improves model robustness and generalization.

Visualization confirms the rationality and effectiveness of the approach.

Abstract

This paper focuses on the domain generalization task where domain knowledge is unavailable, and even worse, only samples from a single domain can be utilized during training. Our motivation originates from the recent progresses in deep neural network (DNN) testing, which has shown that maximizing neuron coverage of DNN can help to explore possible defects of DNN (i.e., misclassification). More specifically, by treating the DNN as a program and each neuron as a functional point of the code, during the network training we aim to improve the generalization capability by maximizing the neuron coverage of DNN with the gradient similarity regularization between the original and augmented samples. As such, the decision behavior of the DNN is optimized, avoiding the arbitrary neurons that are deleterious for the unseen samples, and leading to the trained DNN that can be better generalized to out-of-distribution samples. Extensive studies on various domain generalization tasks based on both single and multiple domain(s) setting demonstrate the effectiveness of our proposed approach compared with state-of-the-art baseline methods. We also analyze our method by conducting visualization based on network dissection. The results further provide useful evidence on the rationality and effectiveness of our approach.