Abstract
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 data sets 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.
Original language | English |
---|---|
Title of host publication | 2019 IEEE Winter Conference on Applications of Computer Vision (WACV) |
Place of Publication | Piscataway |
Publisher | Institute of Electrical and Electronics Engineers |
Pages | 1866-1875 |
Number of pages | 10 |
ISBN (Electronic) | 978-1-7281-1975-5 |
DOIs | |
Publication status | Published - 4 Mar 2019 |
Event | 2019 IEEE Winter Conference on Applications of Computer Vision (WACV) - Waikoloa Village, United States Duration: 7 Jan 2019 → 11 Jan 2019 |
Conference
Conference | 2019 IEEE Winter Conference on Applications of Computer Vision (WACV) |
---|---|
Country | United States |
City | Waikoloa Village |
Period | 7/01/19 → 11/01/19 |