Semisupervised Semantic Segmentation With Certainty-Aware Consistency Training for Remote Sensing Imagery

Abstract

Semi-supervised learning is a forcible method to lessen the cost of annotation for remote sensing semantic segmentation tasks. Recent related researches indicate that consistency training is one of the most effective strategies in semisupervised learning. The core of consistency training is maintaining model outputs consistent under various perturbations. However, the current consistency training-based semi-supervised semantic segmentation frameworks lack the analysis of model uncertainty, which increases the generation of semantic ambiguity on remote sensing images. Therefore, we propose the certainty-aware consistency training (CACT) strategy to mitigate the influence of semantic ambiguity caused by model uncertainty. The certainty-aware consistency training strategy consists of two novel parts: certainty-aware consistency correction (CACC) and class-balanced adaptive threshold (CBAT). The CACC starts with generating a high-quality prediction target, then models the importance of the consistent output target and corrects the output predictions according to the certainty map, increasing the focus on reliable predictions. The CBAT uses a dynamic class-balanced adaptive threshold to filter out unreliable predictions, further reducing the impact of semantic ambiguity. Finally, considerable experimental results on the DLRSD, WHDLD, and Potsdam demonstrate that our framework has excellent performance on semi-supervised remote sensing semantic segmentation scenarios.

Spatial-Logic-Aware Weakly Supervised Learning for Flood Mapping on Earth Imagery

“…Numerous works exist that address weak (e.g., coarse) labels in earth imagery segmentation for convolutional neural networks or vision transformers. Techniques include utilizing teacher-student framework (Wang et al 2020a;Guo et al 2023), designing robust loss function (Mnih and Hinton 2012;Malkin et al 2018), incorporating geometric properties of spatial labels into learning framework (He et al 2022a;Jiang et al 2022), and learning multi-scale features (Yang et al 2012;Robinson et al 2019;Cao and Huang 2022). However, these methods do not incorporate physical knowledge and thus may produce physically implausible results, i.e., erroneous predictions that violate physical constraints.…”