---
title: "Screening, Rectifying, and Re-Screening: A Unified Framework for Tuning Vision-Language Models with Noisy Labels"
authors:
  - "Chaowei Fang"
  - "Hangfei Ma"
  - "Zhihao Li"
  - "De Cheng"
  - "Yue Zhang"
  - "Guanbin Li"
doi: "10.24963/ijcai.2025/568"
url: "https://ijcai.org/proceedings/2025/568"
pdf: "https://ijcai.org/proceedings/2025/0568.pdf"
venue: "Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence"
year: 2025
pages: "5101-5109"
track: "Main Track"
keywords:
  - "Machine Learning: ML: Cost-sensitive learning"
  - "Computer Vision: CV: Machine learning for vision"
  - "Machine Learning: ML: Classification"
  - "Computer Vision: CV: Transfer, low-shot, semi- and un- supervised learning"
---

# Screening, Rectifying, and Re-Screening: A Unified Framework for Tuning Vision-Language Models with Noisy Labels

Chaowei Fang, Hangfei Ma, Zhihao Li, De Cheng, Yue Zhang, Guanbin Li

*Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence*, pages 5101-5109. DOI: [10.24963/ijcai.2025/568](https://doi.org/10.24963/ijcai.2025/568)

## Abstract

Pre-trained vision-language models have shown remarkable potential for downstream tasks. However, their fine-tuning under noisy labels remains an open problem due to challenges like self-confirmation bias and the limitations of conventional small-loss criteria. In this paper, we propose a unified framework to address these issues, consisting of three key steps: Screening, Rectifying, and Re-Screening. First, a dual-level semantic matching mechanism is introduced to categorize samples into clean, ambiguous, and noisy samples by leveraging both macro-level and micro-level textual prompts. Second, we design tailored pseudo-labeling strategies to rectify noisy and ambiguous labels, enabling their effective incorporation into the training process. Finally, a re-screening step, utilizing cross-validation with an auxiliary vision-language model, mitigates self-confirmation bias and enhances the robustness of the framework. Extensive experiments across ten datasets demonstrate that the proposed method significantly outperforms existing approaches for tuning vision-language pre-trained models with noisy labels.