Bridging the Visual Gap: Fine-Tuning Multimodal Models with Knowledge-Adapted Captions

November 13, 2024 Β· Declared Dead Β· πŸ› North American Chapter of the Association for Computational Linguistics

πŸ‘» CAUSE OF DEATH: Ghosted
No code link whatsoever

"No code URL or promise found in abstract"

Evidence collected by the PWNC Scanner

Authors Moran Yanuka, Assaf Ben Kish, Yonatan Bitton, Idan Szpektor, Raja Giryes arXiv ID 2411.09018 Category cs.CV: Computer Vision Cross-listed cs.CL, cs.LG Citations 5 Venue North American Chapter of the Association for Computational Linguistics Last Checked 4 months ago
Abstract
Recent research increasingly focuses on training vision-language models (VLMs) with long, detailed image captions. However, small-scale VLMs often struggle to balance the richness of these captions with the risk of hallucinating content during fine-tuning. In this paper, we explore how well VLMs adapt to such captions. To quantify caption quality, we propose Decomposed NLI (DNLI), an evaluation framework that breaks down generated captions into individual propositions, assessing each in isolation. This fine-grained analysis reveals a critical balance between capturing descriptive details and preventing hallucinations. Our findings show that simply reducing caption complexity or employing standard data curation techniques does not effectively resolve this issue. To tackle this challenge, we introduce Knowledge Adapted (KnowAda) fine-tuning, a data-centric approach that automatically adapts training data with the model's existing knowledge and visual understanding. KnowAda minimizes hallucinations while preserving high descriptiveness. We validate this approach across several small-scale VLMs (up to 7B parameters) and dense caption datasets, demonstrating that KnowAda effectively balances hallucination reduction and descriptiveness. Our results show that KnowAda outperforms various baselines in both automatic metrics and human evaluations. We will release our code and models.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

πŸ“œ Similar Papers

In the same crypt β€” Computer Vision

πŸŒ… πŸŒ… Old Age

Fast R-CNN

Ross Girshick

cs.CV πŸ› ICCV πŸ“š 27.7K cites 11 years ago

Died the same way β€” πŸ‘» Ghosted