Class-Aware Prototype Learning with Negative Contrast for Test-Time Adaptation of Vision-Language Models

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Qiao, Xiaozhen, Zhao, Jingkai, Jiang, Yuqiu, Guo, Xianda, Sun, Zhe, Zhang, Hongyuan, Li, Xuelong
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912665192890368
author Qiao, Xiaozhen
Zhao, Jingkai
Jiang, Yuqiu
Guo, Xianda
Sun, Zhe
Zhang, Hongyuan
Li, Xuelong
author_facet Qiao, Xiaozhen
Zhao, Jingkai
Jiang, Yuqiu
Guo, Xianda
Sun, Zhe
Zhang, Hongyuan
Li, Xuelong
contents Vision-Language Models (VLMs) demonstrate impressive zero-shot generalization through large-scale image-text pretraining, yet their performance can drop once the deployment distribution diverges from the training distribution. To address this, Test-Time Adaptation (TTA) methods update models using unlabeled target data. However, existing approaches often ignore two key challenges: prototype degradation in long-tailed distributions and confusion between semantically similar classes. To tackle these issues, we propose \textbf{C}lass-Aware \textbf{P}rototype \textbf{L}earning with \textbf{N}egative \textbf{C}ontrast(\textbf{CPL-NC}), a lightweight TTA framework designed specifically for VLMs to enhance generalization under distribution shifts. CPL-NC introduces a \textit{Class-Aware Prototype Cache} Module that dynamically adjusts per-class capacity based on test-time frequency and activation history, with a rejuvenation mechanism for inactive classes to retain rare-category knowledge. Additionally, a \textit{Negative Contrastive Learning} Mechanism identifies and constrains hard visual-textual negatives to improve class separability. The framework employs asymmetric optimization, refining only textual prototypes while anchoring on stable visual features. Experiments on 15 benchmarks show that CPL-NC consistently outperforms prior TTA methods across both ResNet-50 and ViT-B/16 backbones.
format Preprint
id arxiv_https___arxiv_org_abs_2510_19802
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Class-Aware Prototype Learning with Negative Contrast for Test-Time Adaptation of Vision-Language Models
Qiao, Xiaozhen
Zhao, Jingkai
Jiang, Yuqiu
Guo, Xianda
Sun, Zhe
Zhang, Hongyuan
Li, Xuelong
Computer Vision and Pattern Recognition
Vision-Language Models (VLMs) demonstrate impressive zero-shot generalization through large-scale image-text pretraining, yet their performance can drop once the deployment distribution diverges from the training distribution. To address this, Test-Time Adaptation (TTA) methods update models using unlabeled target data. However, existing approaches often ignore two key challenges: prototype degradation in long-tailed distributions and confusion between semantically similar classes. To tackle these issues, we propose \textbf{C}lass-Aware \textbf{P}rototype \textbf{L}earning with \textbf{N}egative \textbf{C}ontrast(\textbf{CPL-NC}), a lightweight TTA framework designed specifically for VLMs to enhance generalization under distribution shifts. CPL-NC introduces a \textit{Class-Aware Prototype Cache} Module that dynamically adjusts per-class capacity based on test-time frequency and activation history, with a rejuvenation mechanism for inactive classes to retain rare-category knowledge. Additionally, a \textit{Negative Contrastive Learning} Mechanism identifies and constrains hard visual-textual negatives to improve class separability. The framework employs asymmetric optimization, refining only textual prototypes while anchoring on stable visual features. Experiments on 15 benchmarks show that CPL-NC consistently outperforms prior TTA methods across both ResNet-50 and ViT-B/16 backbones.
title Class-Aware Prototype Learning with Negative Contrast for Test-Time Adaptation of Vision-Language Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2510.19802