open access

https://hdl.handle.net/2115/92818

Learning intra-domain style-invariant representation for unsupervised domain adaptation of semantic segmentation

英語

Domain adaptation

2022-07-20

Elsevier

Pattern recognition

132

108911

A B S T R A C T In this paper, we aim to tackle the problem of unsupervised domain adaptation (UDA) of semantic seg-mentation and improve the UDA performance with a novel conception of learning intra-domain style -invariant representation. Previous UDA methods focused on reducing the inter-domain inconsistency between the source domain and the target domain. However, due to the different data distributions of the two domains, reducing the inter-domain inconsistency cannot ensure the generalization abil-ity of the trained model in the target domain. Therefore, to improve the UDA performance, we take into consideration the intra-domain diversity of the target domain for the first time in studies on UDA and aim to train the model to generalize well to the diverse intra-domain styles. To achieve this, we propose a self-ensembling method to learn the intra-domain style-invariant representation and we in-troduce a semantic-aware multimodal image-to-image translation model to obtain images with diver-sified intra-domain styles. Our method achieves state-of-the-art performance on two synthetic-to-real adaptation benchmarks, and we demonstrate the effectiveness of our method by conducting extensive experiments. (c) 2022 Elsevier Ltd. All rights reserved.

学術雑誌論文

Accepted Manuscript

Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

© 2022. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/

0031-3203

AA00770025