{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,8]],"date-time":"2026-05-08T17:33:00Z","timestamp":1778261580758,"version":"3.51.4"},"reference-count":118,"publisher":"Springer Science and Business Media LLC","issue":"3","license":[{"start":{"date-parts":[[2024,9,15]],"date-time":"2024-09-15T00:00:00Z","timestamp":1726358400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2024,9,15]],"date-time":"2024-09-15T00:00:00Z","timestamp":1726358400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100000923","name":"Australian Research Council","doi-asserted-by":"publisher","award":["DE240100105"],"award-info":[{"award-number":["DE240100105"]}],"id":[{"id":"10.13039\/501100000923","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100000923","name":"Australian Research Council","doi-asserted-by":"publisher","award":["DP240101814"],"award-info":[{"award-number":["DP240101814"]}],"id":[{"id":"10.13039\/501100000923","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Int J Comput Vis"],"published-print":{"date-parts":[[2025,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n          <jats:p>This article presents a comprehensive survey of online test-time adaptation (OTTA), focusing on effectively adapting machine learning models to distributionally different target data upon batch arrival. Despite the recent proliferation of OTTA methods, conclusions from previous studies are inconsistent due to ambiguous settings, outdated backbones, and inconsistent hyperparameter tuning, which obscure core challenges and hinder reproducibility. To enhance clarity and enable rigorous comparison, we classify OTTA techniques into three primary categories and benchmark them using a modern backbone, the Vision Transformer. Our benchmarks cover conventional corrupted datasets such as CIFAR-10\/100-C and ImageNet-C, as well as real-world shifts represented by CIFAR-10.1, OfficeHome, and CIFAR-10-Warehouse. The CIFAR-10-Warehouse dataset includes a variety of variations from different search engines and synthesized data generated through diffusion models. To measure efficiency in online scenarios, we introduce novel evaluation metrics, including GFLOPs, wall clock time, and GPU memory usage, providing a clearer picture of the trade-offs between adaptation accuracy and computational overhead. Our findings diverge from existing literature, revealing that (1) transformers demonstrate heightened resilience to diverse domain shifts, (2) the efficacy of many OTTA methods relies on large batch sizes, and (3) stability in optimization and resistance to perturbations are crucial during adaptation, particularly when the batch size is 1. Based on these insights, we highlight promising directions for future research. Our benchmarking toolkit and source code are available at <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"https:\/\/github.com\/Jo-wang\/OTTA_ViT_survey\" ext-link-type=\"uri\">https:\/\/github.com\/Jo-wang\/OTTA_ViT_survey<\/jats:ext-link>.<\/jats:p>","DOI":"10.1007\/s11263-024-02213-5","type":"journal-article","created":{"date-parts":[[2024,9,15]],"date-time":"2024-09-15T09:02:02Z","timestamp":1726390922000},"page":"1106-1139","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":13,"title":["In Search of Lost Online Test-Time Adaptation: A Survey"],"prefix":"10.1007","volume":"133","author":[{"given":"Zixin","family":"Wang","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6272-2971","authenticated-orcid":false,"given":"Yadan","family":"Luo","sequence":"additional","affiliation":[]},{"given":"Liang","family":"Zheng","sequence":"additional","affiliation":[]},{"given":"Zhuoxiao","family":"Chen","sequence":"additional","affiliation":[]},{"given":"Sen","family":"Wang","sequence":"additional","affiliation":[]},{"given":"Zi","family":"Huang","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2024,9,15]]},"reference":[{"key":"2213_CR1","doi-asserted-by":"publisher","unstructured":"Adachi, K., Yamaguchi, S., & Kumagai, A. (2023). Covariance-aware feature alignment with pre-computed source statistics for test-time adaptation to multiple image corruptions. In ICIP (pp. 800\u2013804). https:\/\/doi.org\/10.1109\/ICIP49359.2023.10222901","DOI":"10.1109\/ICIP49359.2023.10222901"},{"key":"2213_CR2","unstructured":"Ba, L.\u00a0J., Kiros, J.\u00a0R., & Hinton, G.\u00a0E. (2016). Layer normalization. CoRR arXiv:1607.06450"},{"key":"2213_CR3","doi-asserted-by":"publisher","unstructured":"Barron, J.\u00a0T. (2019). A general and adaptive robust loss function. In CVPR (pp. 4331\u20134339). https:\/\/doi.org\/10.1109\/CVPR.2019.00446","DOI":"10.1109\/CVPR.2019.00446"},{"key":"2213_CR4","doi-asserted-by":"publisher","unstructured":"Boudiaf, M., M\u00fcller, R., Ayed, I.\u00a0B., & Bertinetto, L. (2022). Parameter-free online test-time adaptation. In CVPR, pp. 8334\u20138343. https:\/\/doi.org\/10.1109\/CVPR52688.2022.00816","DOI":"10.1109\/CVPR52688.2022.00816"},{"key":"2213_CR5","doi-asserted-by":"publisher","unstructured":"Brahma, D., & Rai, P. (2023). A probabilistic framework for lifelong test-time adaptation. In CVPR (pp. 3582\u20133591). https:\/\/doi.org\/10.1109\/CVPR52729.2023.00349","DOI":"10.1109\/CVPR52729.2023.00349"},{"key":"2213_CR6","doi-asserted-by":"publisher","unstructured":"Carlucci, F.\u00a0M., Porzi, L., Caputo, B., Ricci, E., & Bul\u00f2, S.\u00a0R. (2017). Autodial: Automatic domain alignment layers. In ICCV (pp. 5077\u20135085). https:\/\/doi.org\/10.1109\/ICCV.2017.542","DOI":"10.1109\/ICCV.2017.542"},{"key":"2213_CR7","unstructured":"Chakrabarty, G., Sreenivas, M., & Biswas, S. (2023). SATA: Source anchoring and target alignment network for continual test time adaptation. CoRR arXiv:2304.10113"},{"key":"2213_CR8","doi-asserted-by":"publisher","unstructured":"Chen, Z., Luo, Y., & Baktashmotlagh, M. (2021). Conditional extreme value theory for open set video domain adaptation. In MMAsia. (pp. 20:1\u201320:8). ACM, https:\/\/doi.org\/10.1145\/3469877.3490600","DOI":"10.1145\/3469877.3490600"},{"key":"2213_CR9","doi-asserted-by":"publisher","unstructured":"Chen, Z., Luo, Y., Wang, Z., Baktashmotlagh, M., & Huang, Z. (2023a). Revisiting domain-adaptive 3d object detection by reliable, diverse and class-balanced pseudo-labeling. In ICCV (pp. 3691\u20133703). https:\/\/doi.org\/10.1109\/ICCV51070.2023.00344","DOI":"10.1109\/ICCV51070.2023.00344"},{"key":"2213_CR10","unstructured":"Chen, Z., Luo, Y., Wang, Z., Wang, Z., Yu, X., & Huang, Z. (2023b). Towards open world active learning for 3d object detection. CoRR arXiv:2310.10391"},{"key":"2213_CR11","doi-asserted-by":"publisher","unstructured":"Chen, D., Wang, D., Darrell, T., & Ebrahimi, S. (2022). Contrastive test-time adaptation. In CVPR (pp. 295\u2013305). https:\/\/doi.org\/10.1109\/CVPR52688.2022.00039","DOI":"10.1109\/CVPR52688.2022.00039"},{"key":"2213_CR12","doi-asserted-by":"crossref","unstructured":"Chen, Z., Wang, Z., Wang, S., Huang, Z., & Luo, Y. (2024). DPO: Dual-perturbation optimization for test-time adaptation in 3d object detection. ACM MM.","DOI":"10.1145\/3664647.3681040"},{"key":"2213_CR13","first-page":"747","volume":"10","author":"Y Chen","year":"2009","unstructured":"Chen, Y., Garcia, E. K., Gupta, M. R., Rahimi, A., & Cazzanti, L. (2009). Similarity-based classification: Concepts and algorithms. Journal of Machine Learning Research, 10, 747\u2013776.","journal-title":"Journal of Machine Learning Research"},{"issue":"12","key":"2213_CR14","doi-asserted-by":"publisher","first-page":"9615","DOI":"10.1109\/TPAMI.2021.3129819","volume":"44","author":"M Choi","year":"2021","unstructured":"Choi, M., Choi, J., Baik, S., Kim, T. H., & Lee, K. M. (2021). Test-time adaptation for video frame interpolation via meta-learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(12), 9615\u20139628.","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"2213_CR15","doi-asserted-by":"publisher","first-page":"440","DOI":"10.1007\/978-3-031-19827-4_26","volume":"13693","author":"S Choi","year":"2022","unstructured":"Choi, S., Yang, S., Choi, S., & Yun, S. (2022). Improving test-time adaptation via shift-agnostic weight regularization and nearest source prototypes. ECCV, Lecture Notes in Computer Science, 13693, 440\u2013458. https:\/\/doi.org\/10.1007\/978-3-031-19827-4_26","journal-title":"ECCV, Lecture Notes in Computer Science"},{"key":"2213_CR16","unstructured":"Cogswell, M., Ahmed, F., Girshick, R.\u00a0B., Zitnick, L., & Batra, D. (2016). Reducing overfitting in deep networks by decorrelating representations. In ICLR."},{"key":"2213_CR17","unstructured":"Croce, F., Andriushchenko, M., Sehwag, V., Debenedetti, E., Flammarion, N., Chiang, M., Mittal, P., & Hein, M. (2021). Robustbench: A standardized adversarial robustness benchmark. In NeurIPS benchmarks track."},{"key":"2213_CR18","doi-asserted-by":"crossref","unstructured":"Ding, Y., Liang, J., Jiang, B., Zheng, A., & He, R. (2023). Maps: A noise-robust progressive learning approach for source-free domain adaptive keypoint detection. arXiv:2302.04589","DOI":"10.1109\/TCSVT.2023.3294963"},{"key":"2213_CR19","doi-asserted-by":"publisher","unstructured":"Ding, N., Xu, Y., Tang, Y., Xu, C., Wang, Y., & Tao, D. (2022). Source-free domain adaptation via distribution estimation. In CVPR (pp. 7202\u20137212). https:\/\/doi.org\/10.1109\/CVPR52688.2022.00707","DOI":"10.1109\/CVPR52688.2022.00707"},{"key":"2213_CR20","doi-asserted-by":"publisher","unstructured":"D\u00f6bler, M., Marsden, R.\u00a0A., & Yang, B. (2023). Robust mean teacher for continual and gradual test-time adaptation. In CVPR (pp. 7704\u20137714). https:\/\/doi.org\/10.1109\/CVPR52729.2023.00744","DOI":"10.1109\/CVPR52729.2023.00744"},{"key":"2213_CR21","unstructured":"Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., & Houlsby, N. (2021). An image is worth $$16\\times 16$$ words: Transformers for image recognition at scale. In ICLR."},{"key":"2213_CR22","unstructured":"Foret, P., Kleiner, A., Mobahi, H., & Neyshabur, B. (2021). Sharpness-aware minimization for efficiently improving generalization. In ICLR."},{"key":"2213_CR23","first-page":"1050","volume":"48","author":"Y Gal","year":"2016","unstructured":"Gal, Y., & Ghahramani, Z. (2016). Dropout as a Bayesian approximation: Representing model uncertainty in deep learning. ICML, JMLR Workshop and Conference Proceedings, 48, 1050\u20131059.","journal-title":"ICML, JMLR Workshop and Conference Proceedings"},{"key":"2213_CR24","doi-asserted-by":"publisher","unstructured":"Gan, Y., Bai, Y., Lou, Y., Ma, X., Zhang, R., Shi, N., & Luo, L. (2023). Decorate the newcomers: Visual domain prompt for continual test time adaptation. In AAAI (pp. 7595\u20137603). https:\/\/doi.org\/10.1609\/AAAI.V37I6.25922","DOI":"10.1609\/AAAI.V37I6.25922"},{"key":"2213_CR25","unstructured":"Gandelsman, Y., Sun, Y., Chen, X., & Efros, A.\u00a0A. (2022). Test-time training with masked autoencoders. In NeurIPS."},{"key":"2213_CR26","first-page":"1180","volume":"37","author":"Y Ganin","year":"2015","unstructured":"Ganin, Y., & Lempitsky, V. S. (2015). Unsupervised domain adaptation by backpropagation. ICML, JMLR Workshop and Conference Proceedings, 37, 1180\u20131189.","journal-title":"ICML, JMLR Workshop and Conference Proceedings"},{"key":"2213_CR27","unstructured":"Gao, Y., Shi, X., Zhu, Y., Wang, H., Tang, Z., Zhou, X., Li, M., & Metaxas, D.\u00a0N. (2022). Visual prompt tuning for test-time domain adaptation. CoRR arXiv:2210.04831"},{"key":"2213_CR28","doi-asserted-by":"publisher","unstructured":"Gao, J., Zhang, J., Liu, X., Darrell, T., Shelhamer, E., & Wang, D. (2023). Back to the source: Diffusion-driven adaptation to test-time corruption. In CVPR (pp. 11786\u201311796). https:\/\/doi.org\/10.1109\/CVPR52729.2023.01134","DOI":"10.1109\/CVPR52729.2023.01134"},{"key":"2213_CR29","unstructured":"Gong, T., Jeong, J., Kim, T., Kim, Y., Shin, J., & Lee, S. (2022). NOTE: Robust continual test-time adaptation against temporal correlation. In NeurIPS."},{"key":"2213_CR30","unstructured":"Gong, T., Kim, Y., Lee, T., Chottananurak, S., & Lee, S. (2023). Sotta: Robust test-time adaptation on noisy data streams. CoRR arXiv:2310.10074"},{"key":"2213_CR31","unstructured":"Goyal, S., Sun, M., Raghunathan, A., & Kolter, J.\u00a0Z. (2022). Test time adaptation via conjugate pseudo-labels. In NeurIPS."},{"key":"2213_CR32","doi-asserted-by":"crossref","unstructured":"Han, J., Zeng, L., Du, L., Ding, W., & Feng, J. (2023). Rethinking precision of pseudo label: Test-time adaptation via complementary learning. CoRR arXiv:2301.06013","DOI":"10.1016\/j.patrec.2023.12.001"},{"key":"2213_CR33","doi-asserted-by":"publisher","unstructured":"He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In CVPR (pp. 770\u2013778). https:\/\/doi.org\/10.1109\/CVPR.2016.90","DOI":"10.1109\/CVPR.2016.90"},{"key":"2213_CR34","unstructured":"Hegde, D., Sindagi, V., Kilic, V., Cooper, A.\u00a0B., Foster, M., & Patel, V. (2021). Uncertainty-aware mean teacher for source-free unsupervised domain adaptive 3d object detection. arXiv:2109.14651"},{"key":"2213_CR35","unstructured":"Hendrycks, D., & Dietterich, T.\u00a0G. (2018). Benchmarking neural network robustness to common corruptions and surface variations. arXiv:1807.01697"},{"key":"2213_CR36","unstructured":"Hendrycks, D., Mu, N., Cubuk, E.\u00a0D., Zoph, B., Gilmer, J., & Lakshminarayanan, B. (2020). Augmix: A simple data processing method to improve robustness and uncertainty. In ICLR."},{"key":"2213_CR37","unstructured":"Hong, J., Lyu, L., Zhou, J., & Spranger, M. (2023). MECTA: Memory-economic continual test-time model adaptation. In ICLR."},{"key":"2213_CR38","unstructured":"Hu, X., Uzunbas, M.\u00a0G., Chen, S., Wang, R., Shah, A., Nevatia, R., & Lim, S. (2021). Mixnorm: Test-time adaptation through online normalization estimation. CoRR arXiv:2110.11478"},{"key":"2213_CR39","doi-asserted-by":"crossref","unstructured":"Huang, J., Smola, A.\u00a0J., Gretton, A., Borgwardt, K.\u00a0M., & Sch\u00f6lkopf, B. (2006). Correcting sample selection bias by unlabeled data. In NIPS (pp. 601\u2013608).","DOI":"10.7551\/mitpress\/7503.003.0080"},{"key":"2213_CR40","doi-asserted-by":"publisher","unstructured":"Huang, L., Yang, D., Lang, B., & Deng, J. (2018). Decorrelated batch normalization. In CVPR (pp. 791\u2013800). https:\/\/doi.org\/10.1109\/CVPR.2018.00089","DOI":"10.1109\/CVPR.2018.00089"},{"key":"2213_CR41","unstructured":"Ioffe, S. (2017). Batch renormalization: Towards reducing minibatch dependence in batch-normalized models. In NIPS (pp. 1945\u20131953)."},{"key":"2213_CR42","first-page":"448","volume":"37","author":"S Ioffe","year":"2015","unstructured":"Ioffe, S., & Szegedy, C. (2015). Batch normalization: Accelerating deep network training by reducing internal covariate shift. ICML, JMLR Workshop and Conference Proceedings, 37, 448\u2013456.","journal-title":"ICML, JMLR Workshop and Conference Proceedings"},{"key":"2213_CR43","unstructured":"Iwasawa, Y., & Matsuo, Y. (2021). Test-time classifier adjustment module for model-agnostic domain generalization. In NeurIPS (pp. 2427\u20132440)."},{"key":"2213_CR44","unstructured":"Jang, M., Chung, S., & Chung, H.\u00a0W. (2023). Test-time adaptation via self-training with nearest neighbor information. In ICLR."},{"key":"2213_CR45","doi-asserted-by":"crossref","unstructured":"Jung, S., Lee, J., Kim, N., & Choo, J. (2022). CAFA: Class-aware feature alignment for test-time adaptation. CoRR arXiv:2206.00205","DOI":"10.1109\/ICCV51070.2023.01747"},{"key":"2213_CR46","doi-asserted-by":"publisher","first-page":"558","DOI":"10.1007\/978-3-030-92185-9_46","volume":"13108","author":"M Kimura","year":"2021","unstructured":"Kimura, M. (2021). Understanding test-time augmentation. ICONIP, Lecture Notes in Computer Science, 13108, 558\u2013569. https:\/\/doi.org\/10.1007\/978-3-030-92185-9_46","journal-title":"ICONIP, Lecture Notes in Computer Science"},{"key":"2213_CR47","doi-asserted-by":"publisher","unstructured":"Kingetsu H, Kobayashi K, Okawa Y, Yokota Y, Nakazawa K (2022) Multi-step test-time adaptation with entropy minimization and pseudo-labeling. In ICIP (pp. 4153\u20134157). https:\/\/doi.org\/10.1109\/ICIP46576.2022.9897419","DOI":"10.1109\/ICIP46576.2022.9897419"},{"key":"2213_CR48","doi-asserted-by":"publisher","unstructured":"Kojima, T., Matsuo, Y., & Iwasawa, Y. (2022). Robustifying vision transformer without retraining from scratch by test-time class-conditional feature alignment. In IJCAI (pp. 1009\u20131016). https:\/\/doi.org\/10.24963\/IJCAI.2022\/141","DOI":"10.24963\/IJCAI.2022\/141"},{"key":"2213_CR49","unstructured":"Krizhevsky, A., Sutskever, I., & Hinton, G.\u00a0E. (2012). Imagenet classification with deep convolutional neural networks. In NIPS (pp. 1106\u20131114)."},{"key":"2213_CR50","doi-asserted-by":"publisher","unstructured":"Lee, T., Tremblay, J., Blukis, V., Wen, B., Lee, B., Shin, I., Birchfield, S., Kweon, I.\u00a0S., & Yoon, K. (2023). TTA-COPE: Test-time adaptation for category-level object pose estimation. In CVPR (pp. 21285\u201321295). https:\/\/doi.org\/10.1109\/CVPR52729.2023.02039","DOI":"10.1109\/CVPR52729.2023.02039"},{"key":"2213_CR51","doi-asserted-by":"crossref","unstructured":"Li, Y., Wang, N., Shi, J., Liu, J., & Hou, X. (2017). Revisiting batch normalization for practical domain adaptation. In ICLR.","DOI":"10.1016\/j.patcog.2018.03.005"},{"key":"2213_CR52","unstructured":"Liang, J., He, R., & Tan, T. (2023). A comprehensive survey on test-time adaptation under distribution shifts. CoRR arXiv:2303.15361"},{"key":"2213_CR53","first-page":"6028","volume":"119","author":"J Liang","year":"2020","unstructured":"Liang, J., Hu, D., & Feng, J. (2020). Do we really need to access the source data? Source hypothesis transfer for unsupervised domain adaptation. ICML, 119, 6028\u20136039.","journal-title":"ICML"},{"issue":"11","key":"2213_CR54","first-page":"8602","volume":"44","author":"J Liang","year":"2022","unstructured":"Liang, J., Hu, D., Wang, Y., He, R., & Feng, J. (2022). Source data-absent unsupervised domain adaptation through hypothesis transfer and labeling transfer. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(11), 8602\u20138617.","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"2213_CR55","unstructured":"Lim, H., Kim, B., Choo, J., & Choi, S. (2023). TTN: A domain-shift aware batch normalization in test-time adaptation. In ICLR."},{"key":"2213_CR56","unstructured":"Lin, G., Lai, H., Pan, Y., & Yin, J. (2023). Improving entropy-based test-time adaptation from a clustering view. CoRR arXiv:2310.20327"},{"key":"2213_CR57","doi-asserted-by":"crossref","unstructured":"Liu, H., Chi, Z., Yu, Y., Wang, Y., Chen, J., & Tang, J. (2023a). Meta-auxiliary learning for future depth prediction in videos. In WACV (pp. 5756\u20135765).","DOI":"10.1109\/WACV56688.2023.00571"},{"key":"2213_CR58","unstructured":"Liu, J., Yang, S., Jia, P., Lu, M., Guo, Y., Xue, W., & Zhang, S. (2023b). Vida: Homeostatic visual domain adapter for continual test time adaptation. CoRR arXiv:2306.04344"},{"key":"2213_CR59","doi-asserted-by":"publisher","unstructured":"Luo, Y., Huang, Z., Wang, Z., Zhang, Z., Baktashmotlagh, M. (2020). Adversarial bipartite graph learning for video domain adaptation. In MM (pp. 19\u201327). ACM, https:\/\/doi.org\/10.1145\/3394171.3413897","DOI":"10.1145\/3394171.3413897"},{"issue":"9","key":"2213_CR60","doi-asserted-by":"publisher","first-page":"11240","DOI":"10.1109\/TPAMI.2023.3270288","volume":"45","author":"Y Luo","year":"2023","unstructured":"Luo, Y., Wang, Z., Chen, Z., Huang, Z., & Baktashmotlagh, M. (2023). Source-free progressive graph learning for open-set domain adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(9), 11240\u201311255.","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"2213_CR61","doi-asserted-by":"publisher","first-page":"313","DOI":"10.1007\/978-3-031-16437-8_30","volume":"13433","author":"W Ma","year":"2022","unstructured":"Ma, W., Chen, C., Zheng, S., Qin, J., Zhang, H., & Dou, Q. (2022). Test-time adaptation with calibration of medical image classification nets for label distribution shift. MICCAI, Lecture Notes in Computer Science, 13433, 313\u2013323. https:\/\/doi.org\/10.1007\/978-3-031-16437-8_30","journal-title":"MICCAI, Lecture Notes in Computer Science"},{"key":"2213_CR62","unstructured":"Marsden, R.\u00a0A., D\u00f6bler, M., & Yang, B. (2022). Gradual test-time adaptation by self-training and style transfer. CoRR arXiv:2208.07736"},{"key":"2213_CR63","doi-asserted-by":"publisher","unstructured":"Marsden, R.\u00a0A., D\u00f6bler, M., & Yang, B. (2024). Universal test-time adaptation through weight ensembling, diversity weighting, and prior correction. In WACV (pp. 2543\u20132553). https:\/\/doi.org\/10.1109\/WACV57701.2024.00254","DOI":"10.1109\/WACV57701.2024.00254"},{"key":"2213_CR64","doi-asserted-by":"publisher","unstructured":"Mirza, M.\u00a0J., Micorek, J., Possegger, H., & Bischof, H. (2022). The norm must go on: Dynamic unsupervised domain adaptation by normalization. In CVPR (pp. 14745\u201314755). https:\/\/doi.org\/10.1109\/CVPR52688.2022.01435","DOI":"10.1109\/CVPR52688.2022.01435"},{"issue":"4","key":"2213_CR65","doi-asserted-by":"publisher","first-page":"459","DOI":"10.1002\/int.1018","volume":"16","author":"HB Mitchell","year":"2001","unstructured":"Mitchell, H. B., & Schaefer, P. A. (2001). A \u201csoft\u2019\u2019 k-nearest neighbor voting scheme. International Journal of Intelligent Systems, 16(4), 459\u2013468.","journal-title":"International Journal of Intelligent Systems"},{"key":"2213_CR66","doi-asserted-by":"crossref","unstructured":"Mounsaveng, S., Chiaroni, F., Boudiaf, M., Pedersoli, M., & Ayed, I.\u00a0B. (2023). Bag of tricks for fully test-time adaptation. CoRR arXiv:2310.02416","DOI":"10.1109\/WACV57701.2024.00194"},{"key":"2213_CR67","unstructured":"Mummadi, C.\u00a0K., Hutmacher, R., Rambach, K., Levinkov, E., Brox, T., Metzen, J.\u00a0H. (2021). Test-time adaptation to distribution shift by confidence maximization and input transformation. arXiv:2106.14999"},{"key":"2213_CR68","doi-asserted-by":"publisher","unstructured":"Nguyen, A.\u00a0T., Nguyen-Tang, T., Lim, S., & Torr, P.\u00a0H.\u00a0S. (2023). TIPI: Test time adaptation with transformation invariance. In CVPR (pp. 24162\u201324171). https:\/\/doi.org\/10.1109\/CVPR52729.2023.02314","DOI":"10.1109\/CVPR52729.2023.02314"},{"key":"2213_CR69","doi-asserted-by":"crossref","unstructured":"Niloy, F.\u00a0F., Ahmed, S.\u00a0M., Raychaudhuri, D.\u00a0S., Oymak, S., Roy-Chowdhury, A.\u00a0K. (2023). Effective restoration of source knowledge in continual test time adaptation. CoRR arXiv:2311.04991","DOI":"10.1109\/WACV57701.2024.00209"},{"key":"2213_CR70","unstructured":"Niu, S., Wu, J., Zhang, Y., Wen, Z., Chen, Y., Zhao, P., & Tan, M. (2023). Towards stable test-time adaptation in dynamic wild world. In ICLR."},{"key":"2213_CR71","doi-asserted-by":"publisher","unstructured":"Qiao, F., Zhao, L., & Peng, X. (2020). Learning to learn single domain generalization. In CVPR (pp. 12553\u201312562). https:\/\/doi.org\/10.1109\/CVPR42600.2020.01257","DOI":"10.1109\/CVPR42600.2020.01257"},{"key":"2213_CR72","doi-asserted-by":"crossref","unstructured":"Quinonero-Candela, J., Sugiyama, M., Schwaighofer, A., & Lawrence, N.\u00a0D. (2008). Dataset shift in machine learning. MIT Press.","DOI":"10.7551\/mitpress\/9780262170055.001.0001"},{"key":"2213_CR73","first-page":"8748","volume":"139","author":"A Radford","year":"2021","unstructured":"Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., Krueger, G., & Sutskever, I. (2021). Learning transferable visual models from natural language supervision. ICML, Proceedings of Machine Learning Research, 139, 8748\u20138763.","journal-title":"ICML, Proceedings of Machine Learning Research"},{"key":"2213_CR74","unstructured":"Recht, B., Roelofs, R., Schmidt, L., & Shankar, V. (2018). Do CIFAR-10 classifiers generalize to CIFAR-10?. CoRR arXiv:1806.00451"},{"key":"2213_CR75","doi-asserted-by":"publisher","unstructured":"Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2022). High-resolution image synthesis with latent diffusion models. In CVPR (pp. 10674\u201310685). https:\/\/doi.org\/10.1109\/CVPR52688.2022.01042","DOI":"10.1109\/CVPR52688.2022.01042"},{"key":"2213_CR76","doi-asserted-by":"publisher","unstructured":"Roy, S., Siarohin, A., Sangineto, E., Bul\u00f2, S.\u00a0R., Sebe, N., & Ricci, E. (2019). Unsupervised domain adaptation using feature-whitening and consensus loss. In CVPR (pp. 9471\u20139480). https:\/\/doi.org\/10.1109\/CVPR.2019.00970","DOI":"10.1109\/CVPR.2019.00970"},{"key":"2213_CR77","doi-asserted-by":"publisher","first-page":"567","DOI":"10.1007\/978-3-031-19827-4_33","volume":"13693","author":"C Saltori","year":"2022","unstructured":"Saltori, C., Krivosheev, E., Lathuili\u00e8re, S., Sebe, N., Galasso, F., Fiameni, G., Ricci, E., & Poiesi, F. (2022). GIPSO: Geometrically informed propagation for online adaptation in 3d lidar segmentation. ECCV, Lecture Notes in Computer Science, 13693, 567\u2013585. https:\/\/doi.org\/10.1007\/978-3-031-19827-4_33","journal-title":"ECCV, Lecture Notes in Computer Science"},{"issue":"6","key":"2213_CR78","doi-asserted-by":"publisher","first-page":"863","DOI":"10.1162\/neco.1992.4.6.863","volume":"4","author":"J Schmidhuber","year":"1992","unstructured":"Schmidhuber, J. (1992). Learning factorial codes by predictability minimization. Neural Computation, 4(6), 863\u2013879.","journal-title":"Neural Computation"},{"key":"2213_CR79","doi-asserted-by":"crossref","unstructured":"Seto, S., Theobald, B., Danieli, F., Jaitly, N., & Busbridge, D. (2023). REALM: Robust entropy adaptive loss minimization for improved single-sample test-time adaptation. CoRR arXiv:2309.03964","DOI":"10.1109\/WACV57701.2024.00206"},{"key":"2213_CR80","doi-asserted-by":"publisher","unstructured":"Shanmugam, D., Blalock, D.\u00a0W., Balakrishnan, G., Guttag, J.\u00a0V. (2021). Better aggregation in test-time augmentation. In ICCV (pp. 1194\u20131203). https:\/\/doi.org\/10.1109\/ICCV48922.2021.00125","DOI":"10.1109\/ICCV48922.2021.00125"},{"key":"2213_CR81","unstructured":"Shu, M., Nie, W., Huang, D., Yu, Z., Goldstein, T., Anandkumar, A., & Xiao, C. (2022). Test-time prompt tuning for zero-shot generalization in vision-language models. In NeurIPS."},{"key":"2213_CR82","unstructured":"Sivaprasad, P. T., & Fleuret, F. (2021). Test time adaptation through perturbation robustness. CoRR arXiv:2110.10232"},{"key":"2213_CR83","doi-asserted-by":"crossref","unstructured":"Song, J., Lee, J., Kweon, I.\u00a0S., & Choi, S. (2023). Ecotta: Memory-efficient continual test-time adaptation via self-distilled regularization. CoRR arXiv:2303.01904","DOI":"10.1109\/CVPR52729.2023.01147"},{"key":"2213_CR84","unstructured":"Sun, X., Leng, X., Wang, Z., Yang, Y., Huang, Z., & Zheng, L. (2023). CIFAR-10-warehouse: Broad and more realistic testbeds in model generalization analysis. CoRR arXiv:2310.04414"},{"key":"2213_CR85","first-page":"9229","volume":"119","author":"Y Sun","year":"2020","unstructured":"Sun, Y., Wang, X., Liu, Z., Miller, J., Efros, A. A., & Hardt, M. (2020). Test-time training with self-supervision for generalization under distribution shifts. ICML, Proceedings of Machine Learning Research, 119, 9229\u20139248.","journal-title":"ICML, Proceedings of Machine Learning Research"},{"key":"2213_CR86","unstructured":"Tarvainen, A., & Valpola, H. (2017). Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results. In ICLR."},{"key":"2213_CR87","first-page":"1006","volume":"189","author":"K Thopalli","year":"2022","unstructured":"Thopalli, K., Turaga, P. K., & Thiagarajan, J. J. (2022). Domain alignment meets fully test-time adaptation. ACML, Proceedings of Machine Learning Research, 189, 1006\u20131021.","journal-title":"ACML, Proceedings of Machine Learning Research"},{"key":"2213_CR88","doi-asserted-by":"crossref","unstructured":"Tomar, D., Vray, G., Bozorgtabar, B., & Thiran, J. (2023). Tesla: Test-time self-learning with automatic adversarial augmentation. CoRR arXiv:2303.09870","DOI":"10.1109\/CVPR52729.2023.01948"},{"key":"2213_CR89","unstructured":"Ulyanov, D., Vedaldi, A., & Lempitsky, V.\u00a0S. (2016). Instance normalization: The missing ingredient for fast stylization. CoRR arXiv:1607.08022"},{"key":"2213_CR90","unstructured":"van\u00a0den Oord, A., Li, Y., & Vinyals, O. (2018). Representation learning with contrastive predictive coding. CoRR arXiv:1807.03748"},{"key":"2213_CR91","doi-asserted-by":"publisher","unstructured":"Venkateswara, H., Eusebio, J., Chakraborty, S., & Panchanathan, S. (2017). Deep hashing network for unsupervised domain adaptation. In CVPR (pp. 5385\u20135394). https:\/\/doi.org\/10.1109\/CVPR.2017.572","DOI":"10.1109\/CVPR.2017.572"},{"key":"2213_CR92","doi-asserted-by":"publisher","unstructured":"Wang, Q., Fink, O., Gool, L.\u00a0V., Dai, D (2022a) Continual test-time domain adaptation. In CVPR (pp. 7191\u20137201). https:\/\/doi.org\/10.1109\/CVPR52688.2022.00706","DOI":"10.1109\/CVPR52688.2022.00706"},{"key":"2213_CR93","doi-asserted-by":"publisher","unstructured":"Wang, Z., Luo, Y., Chen, Z., Wang, S., & Huang, Z. (2023b). Cal-SFDA: Source-free domain-adaptive semantic segmentation with differentiable expected calibration error. In MM (pp. 1167\u20131178). ACM, https:\/\/doi.org\/10.1145\/3581783.3611808","DOI":"10.1145\/3581783.3611808"},{"key":"2213_CR94","doi-asserted-by":"publisher","unstructured":"Wang, Z., Luo, Y., Huang, Z., & Baktashmotlagh, M. (2020). Prototype-matching graph network for heterogeneous domain adaptation. In MM (pp. 2104\u20132112). ACM, https:\/\/doi.org\/10.1145\/3394171.3413662","DOI":"10.1145\/3394171.3413662"},{"key":"2213_CR95","doi-asserted-by":"publisher","unstructured":"Wang, Z., Luo, Y., Qiu, R., Huang, Z., Baktashmotlagh, M. (2021b). Learning to diversify for single domain generalization. In ICCV (pp. 814\u2013823). https:\/\/doi.org\/10.1109\/ICCV48922.2021.00087","DOI":"10.1109\/ICCV48922.2021.00087"},{"key":"2213_CR96","doi-asserted-by":"publisher","unstructured":"Wang, Z., Luo, Y., Zhang, P., Wang, S., & Huang, Z. (2022b). Discovering domain disentanglement for generalized multi-source domain adaptation. In ICME (pp. 1\u20136). https:\/\/doi.org\/10.1109\/ICME52920.2022.9859733","DOI":"10.1109\/ICME52920.2022.9859733"},{"key":"2213_CR97","doi-asserted-by":"publisher","unstructured":"Wang, Y., Ma, X., Chen, Z., Luo, Y., Yi, J., & Bailey, J. (2019). Symmetric cross entropy for robust learning with noisy labels. In ICCV (pp. 322\u2013330). https:\/\/doi.org\/10.1109\/ICCV.2019.00041","DOI":"10.1109\/ICCV.2019.00041"},{"key":"2213_CR98","unstructured":"Wang, D., Shelhamer, E., Liu, S., Olshausen, B.\u00a0A., Darrell, T. (2021a). Tent: Fully test-time adaptation by entropy minimization. In ICLR."},{"key":"2213_CR99","unstructured":"Wang, Z., Ye, M., Zhu, X., Peng, L., Tian, L., Zhu, Y. (2022c). Metateacher: Coordinating multi-model domain adaptation for medical image classification. In NeurIPS."},{"key":"2213_CR100","doi-asserted-by":"crossref","unstructured":"Wang, S., Zhang, D., Yan, Z., Zhang, J., & Li, R. (2023a). Feature alignment and uniformity for test time adaptation. CoRR arXiv:2303.10902","DOI":"10.1109\/CVPR52729.2023.01920"},{"key":"2213_CR101","doi-asserted-by":"publisher","first-page":"135","DOI":"10.1016\/j.neucom.2018.05.083","volume":"312","author":"M Wang","year":"2018","unstructured":"Wang, M., & Deng, W. (2018). Deep visual domain adaptation: A survey. Neurocomputing, 312, 135\u2013153.","journal-title":"Neurocomputing"},{"key":"2213_CR102","unstructured":"Wen, Y., Tran, D., Ba, J. (2020). Batchensemble: An alternative approach to efficient ensemble and lifelong learning. In ICLR."},{"key":"2213_CR103","unstructured":"Wilson, A.\u00a0C., Roelofs, R., Stern, M., Srebro, N., & Recht, B. (2017). The marginal value of adaptive gradient methods in machine learning. In NIPS (pp. 4148\u20134158)."},{"key":"2213_CR104","unstructured":"Wu, Q., Yue, X., Sangiovanni-Vincentelli, A. (2021). Domain-agnostic test-time adaptation by prototypical training with auxiliary data. In NeurIPS workshop on distribution shifts."},{"issue":"3","key":"2213_CR105","doi-asserted-by":"publisher","first-page":"742","DOI":"10.1007\/s11263-019-01198-w","volume":"128","author":"Y Wu","year":"2020","unstructured":"Wu, Y., & He, K. (2020). Group normalization. International Journal of Computer Vision, 128(3), 742\u2013755.","journal-title":"International Journal of Computer Vision"},{"key":"2213_CR106","doi-asserted-by":"publisher","unstructured":"Xu, Q., Zhang, R., Zhang, Y., Wang, Y., & Tian, Q. (2021). A Fourier-based framework for domain generalization. In CVPR (pp. 14383\u201314392). https:\/\/doi.org\/10.1109\/CVPR46437.2021.01415","DOI":"10.1109\/CVPR46437.2021.01415"},{"key":"2213_CR107","doi-asserted-by":"crossref","unstructured":"Yang, S., Wang, Y., van\u00a0de Weijer, J., Herranz, L., & Jui, S. (2021). Exploiting the intrinsic neighborhood structure for source-free domain adaptation. In NeurIPS (pp. 29393\u201329405).","DOI":"10.1109\/ICCV48922.2021.00885"},{"key":"2213_CR108","doi-asserted-by":"publisher","unstructured":"Yang, H., Zhang, X., Yin, F., & Liu, C. (2018). Robust classification with convolutional prototype learning. In CVPR (pp. 3474\u20133482). https:\/\/doi.org\/10.1109\/CVPR.2018.00366","DOI":"10.1109\/CVPR.2018.00366"},{"key":"2213_CR109","unstructured":"Yang, T., Zhou, S., Wang, Y., Lu, Y., & Zheng, N. (2022). Test-time batch normalization. CoRR arXiv:2205.10210"},{"issue":"1","key":"2213_CR110","doi-asserted-by":"publisher","first-page":"58","DOI":"10.1109\/JPROC.2015.2494219","volume":"104","author":"J Yang","year":"2016","unstructured":"Yang, J., Meng, X., & Mahoney, M. W. (2016). Implementing randomized matrix algorithms in parallel and distributed environments. Proceedings of the IEEE, 104(1), 58\u201392.","journal-title":"Proceedings of the IEEE"},{"key":"2213_CR111","unstructured":"You, F., Li, J., & Zhao, Z. (2021). Test-time batch statistics calibration for covariate shift. CoRR arXiv:2110.04065"},{"key":"2213_CR112","doi-asserted-by":"crossref","unstructured":"Yuan, L., Xie, B., & Li, S. (2023a). Robust test-time adaptation in dynamic scenarios. CoRR arXiv:2303.13899","DOI":"10.1109\/CVPR52729.2023.01528"},{"key":"2213_CR113","doi-asserted-by":"publisher","unstructured":"Yuan, L., Xie, B., & Li, S. (2023b). Robust test-time adaptation in dynamic scenarios. In CVPR2023 (pp. 15922\u201315932). https:\/\/doi.org\/10.1109\/CVPR52729.2023.01528","DOI":"10.1109\/CVPR52729.2023.01528"},{"key":"2213_CR114","doi-asserted-by":"crossref","unstructured":"Zagoruyko, S., & Komodakis, N. (2016). Wide residual networks. In BMVC. BMVA Press.","DOI":"10.5244\/C.30.87"},{"key":"2213_CR115","unstructured":"Zhang, M., Levine, S., & Finn, C. (2022). MEMO: Test time robustness via adaptation and augmentation. In NeurIPS."},{"key":"2213_CR116","unstructured":"Zhao, B., Chen, C., & Xia, S. (2023a). Delta: Degradation-free fully test-time adaptation. In ICLR."},{"key":"2213_CR117","first-page":"42058","volume":"202","author":"H Zhao","year":"2023","unstructured":"Zhao, H., Liu, Y., Alahi, A., & Lin, T. (2023). On pitfalls of test-time adaptation. ICML, Proceedings of Machine Learning Research, 202, 42058\u201342080.","journal-title":"ICML, Proceedings of Machine Learning Research"},{"key":"2213_CR118","unstructured":"Zhou, K., Liu, Z., Qiao, Y., Xiang, T., & Loy, C. C. (2023). Domain generalization: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(4), 4396\u20134415."}],"container-title":["International Journal of Computer Vision"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11263-024-02213-5.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s11263-024-02213-5\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11263-024-02213-5.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,2,24]],"date-time":"2025-02-24T10:04:35Z","timestamp":1740391475000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s11263-024-02213-5"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,9,15]]},"references-count":118,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2025,3]]}},"alternative-id":["2213"],"URL":"https:\/\/doi.org\/10.1007\/s11263-024-02213-5","relation":{},"ISSN":["0920-5691","1573-1405"],"issn-type":[{"value":"0920-5691","type":"print"},{"value":"1573-1405","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,9,15]]},"assertion":[{"value":"14 December 2023","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"15 July 2024","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"15 September 2024","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}}]}}