Free Lunch for Domain Adversarial Training: Environment Label Smoothing
This addresses a specific problem in domain adaptation/generalization for machine learning practitioners, offering an incremental improvement to existing DAT methods.
The paper tackled training instability in Domain Adversarial Training (DAT) caused by over-confident domain discriminators and noisy environment labels, proposing Environment Label Smoothing (ELS) to improve stability and achieve state-of-the-art results on domain generalization/adaptation tasks, especially with noisy labels.
A fundamental challenge for machine learning models is how to generalize learned models for out-of-distribution (OOD) data. Among various approaches, exploiting invariant features by Domain Adversarial Training (DAT) received widespread attention. Despite its success, we observe training instability from DAT, mostly due to over-confident domain discriminator and environment label noise. To address this issue, we proposed Environment Label Smoothing (ELS), which encourages the discriminator to output soft probability, which thus reduces the confidence of the discriminator and alleviates the impact of noisy environment labels. We demonstrate, both experimentally and theoretically, that ELS can improve training stability, local convergence, and robustness to noisy environment labels. By incorporating ELS with DAT methods, we are able to yield state-of-art results on a wide range of domain generalization/adaptation tasks, particularly when the environment labels are highly noisy.