动态决定何时何地重置，防止模型崩溃并保留有用知识。
When and Where to Reset Matters for Long-Term Test-Time Adaptation
- 根据风险动态选择重置时机与位置，避免盲目周期重置。
- 提出重要性感知正则化，恢复重置丢失的关键知识。
- 适合长期测试时自适应场景，尤其在剧烈领域偏移下有效。
当持续的测试时自适应（TTA）长期进行时，模型会累积错误，导致预测趋于少数类别，即模型崩溃现象。现有研究采用周期性完全重置来消除错误，但这种重置与实际崩溃风险无关，且造成已积累知识的灾难性丢失，而这些知识未来可能仍有用。为此，本文提出：(1) 自适应选择性重置（ASR）机制，动态判断何时何地重置；(2) 重要性感知正则化，恢复因重置丢失的关键知识；(3) 实时自适应调整方案，增强在极端领域偏移下的适应能力。在多个长期TTA基准上的实验表明，该方法在挑战性条件下表现优异。代码已公开于 https://github.com/YonseiML/asr。
原文摘要 · Abstract (English)
When continual test-time adaptation (TTA) persists over the long term, errors accumulate in the model and further cause it to predict only a few classes for all inputs, a phenomenon known as model collapse. Recent studies have explored reset strategies that completely erase these accumulated errors. However, their periodic resets lead to suboptimal adaptation, as they occur independently of the actual risk of collapse. Moreover, their full resets cause catastrophic loss of knowledge acquired over time, even though such knowledge could be beneficial in the future. To this end, we propose (1) an Adaptive and Selective Reset (ASR) scheme that dynamically determines when and where to reset, (2) an importance-aware regularizer to recover essential knowledge lost due to reset, and (3) an on-the-fly adaptation adjustment scheme to enhance adaptability under challenging domain shifts. Extensive experiments across long-term TTA benchmarks demonstrate the effectiveness of our approach, particularly under challenging conditions. Our code is available at https://github.com/YonseiML/asr.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。