用于持续测试时自适应的自适应级联网络
Adaptive Cascading Network for Continual Test-Time Adaptation
AI总结：
针对持续测试时自适应中特征器与分类器不匹配、任务干扰及适应慢的问题，提出元学习预训练的级联网络，并引入平均准确率与前向迁移指标，在多类任务上验证优越性。
摘要：
我们研究持续测试时自适应问题，其目标是在测试时将源预训练模型自适应到一系列未标记目标域。现有测试时训练方法存在若干局限：（1）特征提取器与分类器之间不匹配；（2）主任务与自监督任务之间相互干扰；（3）缺乏快速适应当前分布的能力。针对这些挑战，我们提出一种级联范式，在测试时同时更新特征提取器和分类器，缓解二者之间的不匹配，并实现长期模型自适应。我们模型的预训练构建在元学习框架内，从而最小化主任务与自监督任务之间的干扰，并在有限未标记数据存在时促进快速自适应。此外，我们引入创新的评估指标——平均准确率和前向迁移，以有效衡量模型在动态真实场景中的自适应能力。大量实验和消融研究证明，我们的方法在图像分类、文本分类和语音识别等一系列任务中具有优越性。
英文摘要：
We study the problem of continual test-time adaption where the goal is to adapt a source pre-trained model to a sequence of unlabelled target domains at test time. Existing methods on test-time training suffer from several limitations: (1) Mismatch between the feature extractor and classifier; (2) Interference between the main and self-supervised tasks; (3) Lack of the ability to quickly adapt to the current distribution. In light of these challenges, we propose a cascading paradigm that simultaneously updates the feature extractor and classifier at test time, mitigating the mismatch between them and enabling long-term model adaptation. The pre-training of our model is structured within a meta-learning framework, thereby minimizing the interference between the main and self-supervised tasks and encouraging fast adaptation in the presence of limited unlabelled data. Additionally, we introduce innovative evaluation metrics, average accuracy and forward transfer, to effectively measure the model's adaptation capabilities in dynamic, real-world scenarios. Extensive experiments and ablation studies demonstrate the superiority of our approach in a range of tasks including image classification, text classification, and speech recognition.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
