Postdoc, Institute of automation, Chinese academy of science, Chinese Academy of Sciences
2 papers at NeurIPS 2025
Existing TTA methods mainly focus on single-modal domain shifts and often produce suboptimal results under multi-modal domain shifts due to severe prediction bias. To address this, we propose a novel Partition-Then-Adapt (PTA) method.
We propose DAA, a trainable module that enables real-time adaptation by amplifying feature-level discrepancies between known and unknown classes in TTD.
