Lune

CVPR Findings2026

Mitigating the ID–OOD Tradeoff in Open-Set Test-Time Adaptation

Wenjie Zhao, Jia Li, Xin Dong, Yapeng Tian, Yu Xiang, Yunhui Guo

Abstract

Open-set test-time adaptation (OSTTA) addresses the challenge of adapting models to new environments where outof-distribution (OOD) samples coexist with in-distribution (ID) samples affected by distribution shifts. In such settings, covariate shift-for example, changes in weather conditions such as snow-can alter ID samples, reducing model reliability. Consequently, models must not only correctly classify covariate-shifted ID (csID) samples but also effectively reject covariate-shifted OOD (csOOD) samples. Entropy minimization is a common strategy in test-time adaptation to maintain ID performance under distribution shifts, while entropy maximization is widely applied to enhance OOD detection. Several studies have sought to combine these objectives to tackle the challenges of OSTTA. However, the intrinsic conflict between entropy minimization and maximization inevitably leads to a trade-off between csID classification and csOOD detection. In this paper, we first analyze the limitations of entropy maximization in OSTTA and then introduce an angular loss to regulate feature norm magnitudes, along with a feature-norm loss to suppress csOOD logits, thereby improving OOD detection. These objectives form ROSETTA, a robust open-set test-time adaptation method. Our method achieves strong OOD detection while maintaining high ID classification performance on CIFAR-10-C, CIFAR-100-C, Tiny-ImageNet-C and ImageNet-C. Furthermore, experiments on the Cityscapes validate the method's effectiveness in real-world semantic segmentation, and results on the HAC dataset demonstrate its applicability across different open-set TTA setups.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • ACDC: The Adverse Conditions Dataset with Correspondences for Semantic Driving Scene Understanding cites SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers
  • Tent: Fully Test-Time Adaptation by Entropy Minimization cites Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation
  • Tent: Fully Test-Time Adaptation by Entropy Minimization cites AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on23

  1. SegFormer: Simple and Efficient Design for Semantic Segmentation with TransformersEnze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar et al.NeurIPS 2021 · 9,661 citations
  2. Tent: Fully Test-Time Adaptation by Entropy MinimizationDequan Wang, Evan Shelhamer, Shaoteng Liu, Bruno A. Olshausen et al.ICLR 2021 · 1,731 citations
  3. Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain AdaptationJian Liang, Dapeng Hu, Jiashi FengICML 2020 · 1,624 citations
  4. AugMix: A Simple Data Processing Method to Improve Robustness and UncertaintyDan Hendrycks, Norman Mu, Ekin Dogus Cubuk, Barret Zoph et al.ICLR 2020 · 1,572 citations
  5. ACDC: The Adverse Conditions Dataset with Correspondences for Semantic Driving Scene UnderstandingChristos Sakaridis, Dengxin Dai, Luc Van GoolICCV 2021 · 655 citations

Related papers

  1. Unified Entropy Optimization for Open-Set Test-Time AdaptationZhengqing Gao, Xu-Yao Zhang, Cheng-Lin LiuCVPR 2024
  2. CLIPTTA: Robust Contrastive Vision-Language Test-Time AdaptationMarc Lafon, Gustavo Adolfo Vargas Hakim, Clément Rambour, Christian Desrosiers et al.NeurIPS 2025 · 5 citations
  3. MoETTA: Test-Time Adaptation Under Mixed Distribution Shifts with MoE-LayerNormXiao Fan, Jingyan Jiang, Zhaoru Chen, Fanding Huang et al.AAAI 2026
  4. ODS: Test-Time Adaptation in the Presence of Open-World Data ShiftZhi Zhou, Lan-Zhe Guo, Lin-Han Jia, Dingchu Zhang et al.ICML 2023 · 41 citations
  5. Efficient Test-Time Model Adaptation without ForgettingShuaicheng Niu, Jiaxiang Wu, Yifan Zhang, Yaofo Chen et al.ICML 2022 · 579 citations
Dusk over the sea between two cliffs drawn in fine vertical lines