A Comprehensive Survey on Test-Time Adaptation under Distribution Shifts

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Liang, Jian, He, Ran, Tan, Tieniu
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909424902209536
author Liang, Jian
He, Ran
Tan, Tieniu
author_facet Liang, Jian
He, Ran
Tan, Tieniu
contents Machine learning methods strive to acquire a robust model during the training process that can effectively generalize to test samples, even in the presence of distribution shifts. However, these methods often suffer from performance degradation due to unknown test distributions. Test-time adaptation (TTA), an emerging paradigm, has the potential to adapt a pre-trained model to unlabeled data during testing, before making predictions. Recent progress in this paradigm has highlighted the significant benefits of using unlabeled data to train self-adapted models prior to inference. In this survey, we categorize TTA into several distinct groups based on the form of test data, namely, test-time domain adaptation, test-time batch adaptation, and online test-time adaptation. For each category, we provide a comprehensive taxonomy of advanced algorithms and discuss various learning scenarios. Furthermore, we analyze relevant applications of TTA and discuss open challenges and promising areas for future research. For a comprehensive list of TTA methods, kindly refer to \url{https://github.com/tim-learn/awesome-test-time-adaptation}.
format Preprint
id arxiv_https___arxiv_org_abs_2303_15361
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Comprehensive Survey on Test-Time Adaptation under Distribution Shifts
Liang, Jian
He, Ran
Tan, Tieniu
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Machine learning methods strive to acquire a robust model during the training process that can effectively generalize to test samples, even in the presence of distribution shifts. However, these methods often suffer from performance degradation due to unknown test distributions. Test-time adaptation (TTA), an emerging paradigm, has the potential to adapt a pre-trained model to unlabeled data during testing, before making predictions. Recent progress in this paradigm has highlighted the significant benefits of using unlabeled data to train self-adapted models prior to inference. In this survey, we categorize TTA into several distinct groups based on the form of test data, namely, test-time domain adaptation, test-time batch adaptation, and online test-time adaptation. For each category, we provide a comprehensive taxonomy of advanced algorithms and discuss various learning scenarios. Furthermore, we analyze relevant applications of TTA and discuss open challenges and promising areas for future research. For a comprehensive list of TTA methods, kindly refer to \url{https://github.com/tim-learn/awesome-test-time-adaptation}.
title A Comprehensive Survey on Test-Time Adaptation under Distribution Shifts
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2303.15361