Salvato in:
Dettagli Bibliografici
Autori principali: Lei, Yuqing, Du, Yingjun, Huang, Yawen, Zhen, Xiantong, Shao, Ling
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2512.12268
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911316693745664
author Lei, Yuqing
Du, Yingjun
Huang, Yawen
Zhen, Xiantong
Shao, Ling
author_facet Lei, Yuqing
Du, Yingjun
Huang, Yawen
Zhen, Xiantong
Shao, Ling
contents Vision-language models (VLMs) such as CLIP exhibit strong zero-shot generalization but remain sensitive to domain shifts at test time. Test-time prompt tuning (TPT) mitigates this issue by adapting prompts with fixed augmentations, which may falter in more challenging settings. In this work, we propose Meta Test-Time Prompt Tuning (MetaTPT), a meta-learning framework that learns a self-supervised auxiliary task to guide test-time prompt tuning. The auxiliary task dynamically learns parameterized augmentations for each sample, enabling more expressive transformations that capture essential features in target domains. MetaTPT adopts a dual-loop optimization paradigm: an inner loop learns a self-supervised task that generates informative views, while the outer loop performs prompt tuning by enforcing consistency across these views. By coupling augmentation learning with prompt tuning, MetaTPT improves test-time adaptation under domain shifts. Extensive experiments demonstrate that MetaTPT achieves state-of-the-art performance on domain generalization and cross-dataset benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2512_12268
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MetaTPT: Meta Test-time Prompt Tuning for Vision-Language Models
Lei, Yuqing
Du, Yingjun
Huang, Yawen
Zhen, Xiantong
Shao, Ling
Computer Vision and Pattern Recognition
Vision-language models (VLMs) such as CLIP exhibit strong zero-shot generalization but remain sensitive to domain shifts at test time. Test-time prompt tuning (TPT) mitigates this issue by adapting prompts with fixed augmentations, which may falter in more challenging settings. In this work, we propose Meta Test-Time Prompt Tuning (MetaTPT), a meta-learning framework that learns a self-supervised auxiliary task to guide test-time prompt tuning. The auxiliary task dynamically learns parameterized augmentations for each sample, enabling more expressive transformations that capture essential features in target domains. MetaTPT adopts a dual-loop optimization paradigm: an inner loop learns a self-supervised task that generates informative views, while the outer loop performs prompt tuning by enforcing consistency across these views. By coupling augmentation learning with prompt tuning, MetaTPT improves test-time adaptation under domain shifts. Extensive experiments demonstrate that MetaTPT achieves state-of-the-art performance on domain generalization and cross-dataset benchmarks.
title MetaTPT: Meta Test-time Prompt Tuning for Vision-Language Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.12268