Free Lunch for Domain Adversarial Training: Environment Label Smoothing

YiFan Zhang, xue wang, Jian Liang, Zhang Zhang, Liang Wang, Rong Jin, Tieniu Tan

Keywords: Out-of-Distribution Generalization, Domain adaptation/generalization, Domain adversarial training, environmnt label noise, non-asymptotic convergence

TL;DR: We propose to smooth environment label for domain adversarial training methods, which is experimentally and theoretically shown able to improve training stability, local convergence, and robustness to noisy labels.

Abstract: 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.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Deep Learning and representational learning

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/free-lunch-for-domain-adversarial-training/code)