Mitigating Endogenous Confirmation Bias in Noisy Label Learning for Vision-Language Models

Feiyang Ning; Xinyang Chen

AAAI 2026technical0 citations

Mitigating Endogenous Confirmation Bias in Noisy Label Learning for Vision-Language Models

Abstract

Pretrained vision-language models (VLMs), especially CLIP, excel at adapting to downstream tasks through fine-tuning with sufficient high-quality labeled data. However, real-world training data often contains noisy labels, leading to significant performance degradation when models are naively fine-tuned on them. Existing noisy label learning methods for VLMs typically leverage the model

BibTeX

@inproceedings{aaai2026_mitigatingendoge,
  title = {Mitigating Endogenous Confirmation Bias in Noisy Label Learning for Vision-Language Models},
  author = {Feiyang Ning and Xinyang Chen},
  booktitle = {AAAI 2026},
  year = {2026}
}