Significance-aware Information Bottleneck for Domain Adaptive Semantic Segmentation

Abstract

For unsupervised domain adaptation problems, the strategy of aligning the two domains in latent feature space through adversarial learning has achieved much progress in image classification, but usually fails in semantic segmentation tasks in which the latent representations are overcomplex. In this work, we equip the adversarial network with a "significance-aware information bottleneck (SIB)", to address the above problem. The new network structure, called SIBAN, enables a significance-aware feature purification before the adversarial adaptation, which eases the feature alignment and stabilizes the adversarial training course. In two domain adaptation tasks, i.e., GTA5 -> Cityscapes and SYNTHIA -> Cityscapes, we validate that the proposed method can yield leading results compared with other feature-space alternatives. Moreover, SIBAN can even match the state-of-the-art output-space methods in segmentation accuracy, while the latter are often considered to be better choices for domain adaptive segmentation task.

12 pages, 9 figures

References in corpus (6)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation
• CyCADA: Cycle-Consistent Adversarial Domain Adaptation
• Adversarial Discriminative Domain Adaptation
• Dilated Residual Networks
• Playing for Data: Ground Truth from Computer Games

Cited by in corpus (3)

• Adversarial Style Mining for One-Shot Unsupervised Domain Adaptation
• Domain adaptation based self-correction model for COVID-19 infection segmentation in CT images
• Contextual-Relation Consistent Domain Adaptation for Semantic Segmentation