DOTA: Distributional Test-Time Adaptation of Vision-Language Models

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Han, Zongbo, Yang, Jialong, Wang, Guangyu, Li, Junfan, Xu, Qianli, Shou, Mike Zheng, Zhang, Changqing
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909807337799680
author Han, Zongbo
Yang, Jialong
Wang, Guangyu
Li, Junfan
Xu, Qianli
Shou, Mike Zheng
Zhang, Changqing
author_facet Han, Zongbo
Yang, Jialong
Wang, Guangyu
Li, Junfan
Xu, Qianli
Shou, Mike Zheng
Zhang, Changqing
contents Vision-language foundation models (VLMs), such as CLIP, exhibit remarkable performance across a wide range of tasks. However, deploying these models can be unreliable when significant distribution gaps exist between training and test data, while fine-tuning for diverse scenarios is often costly. Cache-based test-time adapters offer an efficient alternative by storing representative test samples to guide subsequent classifications. Yet, these methods typically employ naive cache management with limited capacity, leading to severe catastrophic forgetting when samples are inevitably dropped during updates. In this paper, we propose DOTA (DistributiOnal Test-time Adaptation), a simple yet effective method addressing this limitation. Crucially, instead of merely memorizing individual test samples, DOTA continuously estimates the underlying distribution of the test data stream. Test-time posterior probabilities are then computed using these dynamically estimated distributions via Bayes' theorem for adaptation. This distribution-centric approach enables the model to continually learn and adapt to the deployment environment. Extensive experiments validate that DOTA significantly mitigates forgetting and achieves state-of-the-art performance compared to existing methods.
format Preprint
id arxiv_https___arxiv_org_abs_2409_19375
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DOTA: Distributional Test-Time Adaptation of Vision-Language Models
Han, Zongbo
Yang, Jialong
Wang, Guangyu
Li, Junfan
Xu, Qianli
Shou, Mike Zheng
Zhang, Changqing
Machine Learning
Artificial Intelligence
Computation and Language
Computer Vision and Pattern Recognition
Human-Computer Interaction
Vision-language foundation models (VLMs), such as CLIP, exhibit remarkable performance across a wide range of tasks. However, deploying these models can be unreliable when significant distribution gaps exist between training and test data, while fine-tuning for diverse scenarios is often costly. Cache-based test-time adapters offer an efficient alternative by storing representative test samples to guide subsequent classifications. Yet, these methods typically employ naive cache management with limited capacity, leading to severe catastrophic forgetting when samples are inevitably dropped during updates. In this paper, we propose DOTA (DistributiOnal Test-time Adaptation), a simple yet effective method addressing this limitation. Crucially, instead of merely memorizing individual test samples, DOTA continuously estimates the underlying distribution of the test data stream. Test-time posterior probabilities are then computed using these dynamically estimated distributions via Bayes' theorem for adaptation. This distribution-centric approach enables the model to continually learn and adapt to the deployment environment. Extensive experiments validate that DOTA significantly mitigates forgetting and achieves state-of-the-art performance compared to existing methods.
title DOTA: Distributional Test-Time Adaptation of Vision-Language Models
topic Machine Learning
Artificial Intelligence
Computation and Language
Computer Vision and Pattern Recognition
Human-Computer Interaction
url https://arxiv.org/abs/2409.19375