Adaptive Spatial Augmentation for Semi-supervised Semantic Segmentation

TL;DR

This paper introduces Adaptive Spatial Augmentation (ASAug), a novel method that dynamically applies spatial transformations to enhance semi-supervised semantic segmentation, leading to improved performance and state-of-the-art results.

Contribution

It demonstrates the effectiveness of spatial augmentation in semi-supervised segmentation and proposes an adaptive strategy to optimize augmentation per image, which was previously underexplored.

Findings

ASAug improves existing semi-supervised segmentation methods.

Spatial augmentation contributes significantly despite label inconsistency.

Achieves state-of-the-art results on PASCAL VOC 2012, Cityscapes, and COCO.

Abstract

In semi-supervised semantic segmentation (SSSS), data augmentation plays a crucial role in the weak-to-strong consistency regularization framework, as it enhances diversity and improves model generalization. Recent strong augmentation methods have primarily focused on intensity-based perturbations, which have minimal impact on the semantic masks. In contrast, spatial augmentations like translation and rotation have long been acknowledged for their effectiveness in supervised semantic segmentation tasks, but they are often ignored in SSSS. In this work, we demonstrate that spatial augmentation can also contribute to model training in SSSS, despite generating inconsistent masks between the weak and strong augmentations. Furthermore, recognizing the variability among images, we propose an adaptive augmentation strategy that dynamically adjusts the augmentation for each instance based on entropy. Extensive experiments show that our proposed Adaptive Spatial Augmentation (\textbf{ASAug}) can be integrated as a pluggable module, consistently improving the performance of existing methods and achieving state-of-the-art results on benchmark datasets such as PASCAL VOC 2012, Cityscapes, and COCO.