SECOND: Mitigating Perceptual Hallucination in Vision-Language Models via Selective and Contrastive Decoding

Woohyeon Park; Woojin Kim; Jaeik Kim; Jaeyoung Do

2025 ICML ICML 2025

SECOND: Mitigating Perceptual Hallucination in Vision-Language Models via Selective and Contrastive Decoding

Abstract

No abstract available. No abstract was captured for this paper when its proceedings were scraped - most often an older edition published without one online. How the corpus is built.

Authors

Woohyeon Park, Woojin Kim, Jaeik Kim, Jaeyoung Do

Topics

Artificial Intelligence > Core AI > Multimodal Learning Machine Learning > Core Methods > Representation Learning Deep Learning > Learning Types > Multimodal Learning

Keywords

vision-language model contrastive decoding selective decoding multi-scale visual information hallucinations object-centric interpretation

Download PDF