Contrastive Test-Time Adaptation

By Dian Chen, Dequan Wang, Trevor Darrell, Sayna Ebrahimi. Computer Vision and Pattern Recognition, 2022

401 citations recorded by the source corpus.

Test-time adaptation is a special setting of unsupervised domain adaptation where a trained model on the source domain has to adapt to the target domain without accessing source data. We propose a novel way to leverage self-supervised contrastive learning to facilitate target feature learning, along with an online pseudo labeling scheme with refinement that significantly denoises pseudo labels. The contrastive learning task is applied jointly with pseudo labeling, contrasting positive and negative pairs constructed similarly as MoCo but with source-initialized encoder, and excluding same-class negative pairs indicated by pseudo labels. Meanwhile, we produce pseudo labels online and refine them via soft voting among their nearest neighbors in the target feature space, enabled by maintaining a memory queue. Our method, AdaContrast, achieves state-of-the-art performance on major benchmarks while having several desirable properties compared to existing works, including memory efficiency, insensitivity to hyper-parameters, and better model calibration. Code is released at https://github.com/DianCh/AdaContrast.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/b4362fe4e0cf17ea212448b2ab683a162dc38489

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Mingkui Tan
2. Shuaicheng Niu
3. Yifan Zhang
4. Peilin Zhao
5. Yaofo Chen
6. Peidong Jia
7. Zhiquan Wen
8. Jiaming Liu

In Computer Vision and Pattern Recognition

1. Mingkui Tan
2. Shuaicheng Niu
3. Yifan Zhang
4. Peilin Zhao
5. Yaofo Chen
6. Peidong Jia
7. Zhiquan Wen
8. Jiaming Liu

Citations in the corpus

• A Comprehensive Survey on Test-Time Adaptation Under Distribution Shifts (2023, neutral): “(5), while other TTDA methods [149, 97, 106, 139] replace p ˜ θ ( y | x i ) with hard pseudo labels for target data under weak augmentation, followed by a cross-entropy loss for target data under strong augmentation.”
• Towards Stable Test-Time Adaptation in Dynamic Wild World (2023, background): “…et al., 2021; Lim et al., 2023; Zhao et al., 2023), unsupervised entropy minimization (Wang et al., 2021; Niu et al., 2022a; Zhang et al., 2022), prediction consistency maximization (Zhang et al., 2022; Wang et al., 2022; Chen et al., 2022a), top-k classification boosting (Niu et al., 2022b), etc.”
• ViDA: Homeostatic Visual Domain Adapter for Continual Test Time Adaptation (2023, methodology): “To tackle this problem, Continual Test-Time Adaptation (CTTA) has been proposed (Wang et al., 2022), which addresses a sequence of different distribution shifts over time rather than a single shift as in TTA.”