ClassHyPer: ClassMix-Based Hybrid Perturbations for Deep Semi-Supervised Semantic Segmentation of Remote Sensing Imagery

• Published in: Remote sensing (Basel, Switzerland) Vol. 14; no. 4; p. 879
• Main Authors: He, Yongjun, Wang, Jinfei, Liao, Chunhua, Shan, Bo, Zhou, Xin
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.02.2022
• Subjects: Annotations; Benchmarks; Classification; Confidence; Consistency; consistency regularization; Datasets; Deep learning; hybrid perturbation; Image annotation; Image processing; Image segmentation; Labeling; Labels; Land cover; Methods; Perturbation; Regularization; Remote sensing; remote sensing semantic segmentation; Satellites; Semantic segmentation; Semantics; Semi-supervised learning; Training; transfer learning
• ISSN: 2072-4292; 2072-4292
• DOI: 10.3390/rs14040879

Inspired by the tremendous success of deep learning (DL) and the increased availability of remote sensing data, DL-based image semantic segmentation has attracted growing interest in the remote sensing community. The ideal scenario of DL application requires a vast number of annotation data with the same feature distribution as the area of interest. However, obtaining such enormous training sets that suit the data distribution of the target area is highly time-consuming and costly. Consistency-regularization-based semi-supervised learning (SSL) methods have gained growing popularity thanks to their ease of implementation and remarkable performance. However, there have been limited applications of SSL in remote sensing. This study comprehensively analyzed several advanced SSL methods based on consistency regularization from the perspective of data- and model-level perturbation. Then, an end-to-end SSL approach based on a hybrid perturbation paradigm was introduced to improve the DL model’s performance with a limited number of labels. The proposed method integrates the semantic boundary information to generate more meaningful mixing images when performing data-level perturbation. Additionally, by using implicit pseudo-supervision based on model-level perturbation, it eliminates the need to set extra threshold parameters in training. Furthermore, it can be flexibly paired with the DL model in an end-to-end manner, as opposed to the separated training stages used in the traditional pseudo-labeling. Experimental results for five remote sensing benchmark datasets in the application of segmentation of roads, buildings, and land cover demonstrated the effectiveness and robustness of the proposed approach. It is particularly encouraging that the ratio of accuracy obtained using the proposed method with 5% labels to that using the purely supervised method with 100% labels was more than 89% on all benchmark datasets.