Annotation-Efficient Active Test-Time Adaptation with Conformal Prediction

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Shi, Tingyu, Lyu, Fan, Peng, Shaoliang
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909816463556608
author Shi, Tingyu
Lyu, Fan
Peng, Shaoliang
author_facet Shi, Tingyu
Lyu, Fan
Peng, Shaoliang
contents Active Test-Time Adaptation (ATTA) improves model robustness under domain shift by selectively querying human annotations at deployment, but existing methods use heuristic uncertainty measures and suffer from low data selection efficiency, wasting human annotation budget. We propose Conformal Prediction Active TTA (CPATTA), which first brings principled, coverage-guaranteed uncertainty into ATTA. CPATTA employs smoothed conformal scores with a top-K certainty measure, an online weight-update algorithm driven by pseudo coverage, a domain-shift detector that adapts human supervision, and a staged update scheme balances human-labeled and model-labeled data. Extensive experiments demonstrate that CPATTA consistently outperforms the state-of-the-art ATTA methods by around 5% in accuracy. Our code and datasets are available at https://github.com/tingyushi/CPATTA.
format Preprint
id arxiv_https___arxiv_org_abs_2509_25692
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Annotation-Efficient Active Test-Time Adaptation with Conformal Prediction
Shi, Tingyu
Lyu, Fan
Peng, Shaoliang
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Active Test-Time Adaptation (ATTA) improves model robustness under domain shift by selectively querying human annotations at deployment, but existing methods use heuristic uncertainty measures and suffer from low data selection efficiency, wasting human annotation budget. We propose Conformal Prediction Active TTA (CPATTA), which first brings principled, coverage-guaranteed uncertainty into ATTA. CPATTA employs smoothed conformal scores with a top-K certainty measure, an online weight-update algorithm driven by pseudo coverage, a domain-shift detector that adapts human supervision, and a staged update scheme balances human-labeled and model-labeled data. Extensive experiments demonstrate that CPATTA consistently outperforms the state-of-the-art ATTA methods by around 5% in accuracy. Our code and datasets are available at https://github.com/tingyushi/CPATTA.
title Annotation-Efficient Active Test-Time Adaptation with Conformal Prediction
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.25692