A parameter-efficient model-based method named distribution-aware tuning (DAT) that adaptively selects and updates two small groups of trainable parameters to extract target domain-specific and task-relevant knowledge, effectively addressing issues of error accumulation and catastrophic forgetting during continual adaptation.
Hybrid-TTA dynamically selects between Full-Tuning and Efficient-Tuning strategies to effectively adapt segmentation models to continually changing target environments, leveraging a Masked Image Modeling based Adaptation framework for robust and efficient continual adaptation.