A Domain Agnostic Normalization Layer for Unsupervised Adversarial Domain Adaptation
This addresses the challenge of improving network performance on unlabeled datasets in domain adaptation, particularly for semantic segmentation, though it appears incremental as it modifies an existing normalization approach within UADA.
The authors tackled the problem that conventional normalization layers worsen performance in unsupervised adversarial domain adaptation (UADA) for semantic scene segmentation, and proposed a Domain Agnostic Normalization layer that surpasses state-of-the-art results on the GTA5 to Cityscapes benchmark and improves performance on unseen domains like Apolloscape and Mapillary.
We propose a normalization layer for unsupervised domain adaption in semantic scene segmentation. Normalization layers are known to improve convergence and generalization and are part of many state-of-the-art fully-convolutional neural networks. We show that conventional normalization layers worsen the performance of current Unsupervised Adversarial Domain Adaption (UADA), which is a method to improve network performance on unlabeled datasets and the focus of our research. Therefore, we propose a novel Domain Agnostic Normalization layer and thereby unlock the benefits of normalization layers for unsupervised adversarial domain adaptation. In our evaluation, we adapt from the synthetic GTA5 data set to the real Cityscapes data set, a common benchmark experiment, and surpass the state-of-the-art. As our normalization layer is domain agnostic at test time, we furthermore demonstrate that UADA using Domain Agnostic Normalization improves performance on unseen domains, specifically on Apolloscape and Mapillary.