Fully-test-time adaptation (F-TTA) can mitigate performance loss due to distribution shifts between train and test data (1) without access to the training data, and (2) without knowledge of the model training procedure. In online F-TTA, a pre-trained model is adapted using a stream of test samples by minimizing a self-supervised objective, such as entropy minimization. However, models adapted with online using entropy minimization, are unstable especially in single sample settings, leading to degenerate solutions, and limiting the adoption of TTA inference strategies. Prior works identify noisy, or unreliable, samples as a cause of failure in online F-TTA. One solution is to ignore these samples, which can lead to bias in the update procedure, slow adaptation, and poor generalization. In this work, we present a general framework for improving robustness of F-TTA to these noisy samples, inspired by self-paced learning and robust loss functions. Our proposed approach, Robust Entropy Adaptive Loss Minimization (REALM), achieves better adaptation accuracy than previous approaches throughout the adaptation process on corruptions of CIFAR-10 and ImageNet-1K, demonstrating its effectiveness.

翻译：全测试时自适应（F-TTA）可在以下条件下缓解训练数据与测试数据分布差异导致的性能损失：（1）无需访问训练数据；（2）无需了解模型训练过程。在在线F-TTA中，通过最小化自监督目标（如熵最小化），利用测试样本流对预训练模型进行自适应。然而，在线熵最小化适配的模型在单样本场景下尤其不稳定，易导致退化解，从而限制了TTA推理策略的推广应用。现有研究指出，噪声或不可靠样本是导致在线F-TTA失败的原因之一。忽略这些样本虽可作为一种解决方案，却可能导致更新过程产生偏差、自适应速度缓慢及泛化能力下降。本文提出一种受自步学习与鲁棒损失函数启发的通用框架，用于提升F-TTA对噪声样本的鲁棒性。我们提出的方法——鲁棒熵自适应损失最小化（REALM），在CIFAR-10和ImageNet-1K的各类损坏场景下，整个自适应过程中的精度均优于现有方法，验证了其有效性。