IDA: Informed Domain Adaptive Semantic Segmentation

Mixup-based data augmentation has been validated to be a critical stage in the self-training framework for unsupervised domain adaptive semantic segmentation (UDASS), which aims to transfer knowledge from a well-annotated (source) domain to an unlabeled (target) domain. Existing self-training methods usually adopt the popular region-based mixup techniques with a random sampling strategy, which unfortunately ignores the dynamic evolution of different semantics across various domains as training proceeds. To improve the UDA-SS performance, we propose an Informed Domain Adaptation (IDA) model, a self-training framework that mixes the data based on class-level segmentation performance, which aims to emphasize small-region semantics during mixup. In our IDA model, the class-level performance is tracked by an expected confidence score (ECS). We then use a dynamic schedule to determine the mixing ratio for data in different domains. Extensive experimental results reveal that our proposed method is able to outperform the state-of-the-art UDA-SS method by a margin of 1. 1 mIoU in the adaptation of GTA-V to Cityscapes and of 0. 9 mIoU in the adaptation of SYNTHIA to Cityscapes. Code link: https://github.com/ArlenCHEN/IDA.git

Authors

• Zheng Chen 0016
• Zhengming Ding
• Jason M. Gregory
• Lantao Liu

Keywords

• Training
• Adaptation models
• Schedules
• Codes
• Semantic segmentation
• Semantics
• Dynamic scheduling
• Domain Adaptation
• Adaptive Segmentation
• Domain Adaptation For Semantic Segmentation
• Random Sampling
• Data Augmentation
• Target Domain
• Segmentation Performance
• Source Domain
• Maximum And Minimum
• Dynamic Changes
• Basis Functions
• Deep Neural Network
• Classification Performance
• Type Classification
• Teacher Model
• Changes In Data
• Labeled Data
• Source Images
• Pseudo Labels
• Classifier Selection
• Progressive Training
• Label Space
• Student Model
• Source Labels
• Semantic Annotation
• Mixed Data
• Selectivity Ratio
• Target Data