Unsupervised Intra-domain Adaptation for Semantic Segmentation through Self-Supervision

TL;DR

This paper introduces a two-step self-supervised domain adaptation method for semantic segmentation that reduces both inter-domain and intra-domain gaps, improving transfer from synthetic to real images.

Contribution

It proposes a novel approach combining inter-domain adaptation with intra-domain self-supervision to better handle distribution gaps within target data.

Findings

Outperforms existing state-of-the-art methods on benchmark datasets.

Effectively reduces intra-domain variability in target data.

Demonstrates significant improvement in segmentation accuracy.

Abstract

Convolutional neural network-based approaches have achieved remarkable progress in semantic segmentation. However, these approaches heavily rely on annotated data which are labor intensive. To cope with this limitation, automatically annotated data generated from graphic engines are used to train segmentation models. However, the models trained from synthetic data are difficult to transfer to real images. To tackle this issue, previous works have considered directly adapting models from the source data to the unlabeled target data (to reduce the inter-domain gap). Nonetheless, these techniques do not consider the large distribution gap among the target data itself (intra-domain gap). In this work, we propose a two-step self-supervised domain adaptation approach to minimize the inter-domain and intra-domain gap together. First, we conduct the inter-domain adaptation of the model; from this adaptation, we separate the target domain into an easy and hard split using an entropy-based ranking function. Finally, to decrease the intra-domain gap, we propose to employ a self-supervised adaptation technique from the easy to the hard split. Experimental results on numerous benchmark datasets highlight the effectiveness of our method against existing state-of-the-art approaches. The source code is available at https://github.com/feipan664/IntraDA.git.