Authors:
Stephan Brehm
;
Sebastian Scherer
and
Rainer Lienhart
Affiliation:
Department of Computer Science, University of Augsburg, Universitätsstr. 6a, Augsburg, Germany
Keyword(s):
Image Translation, Semi-supervised Learning, Unsupervised Learning, Domain Adaptation, Semantic Segmentation, Synthetic Data, Semantic Consistency, Generative Adversarial Networks.
Abstract:
Unsupervised Domain Adaptation (UDA) aims to adapt models trained on a source domain to a new target domain where no labelled data is available. In this work, we investigate the problem of UDA from a synthetic computer-generated domain to a similar but real-world domain for learning semantic segmentation. We propose a semantically consistent image-to-image translation method in combination with a consistency regularisation method for UDA. We overcome previous limitations on transferring synthetic images to real looking images. We leverage pseudo-labels in order to learn a generative image-to-image translation model that receives additional feedback from semantic labels on both domains. Our method outperforms state-of-the-art methods that combine image-to-image translation and semi-supervised learning on relevant domain adaptation benchmarks, i.e., on GTA5 to Cityscapes and SYNTHIA to Cityscapes.