AgenTeeth: A Model-Agnostic Framework for Suppressing Hallucination in Frozen Vision-Language Models on Dental X-Rays via Tool Evidence Injection
A new framework improves dental AI diagnostic accuracy by integrating tool-based evidence to bypass visual hallucinations.
Vision-language models (VLMs) remain largely unreliable on panoramic dental radiographs and can rely on learned anatomical priors rather than evidence in the image. This is particularly problematic for tooth localization and spatial reasoning, and fine-tuned dental VLMs can retai...
ArXiv CS.CV
Source
Generated insight by OnAirToday AI
AgenTeeth addresses the critical challenge of model hallucination in medical imaging by forcing AI to rely on external evidence rather than potentially misleading anatomical priors. This framework bridges the reliability gap for high-stakes dental diagnostics, paving the way for safer autonomous clinical support tools.
Comments
Keep comments relevant and respectful. No links, advertising, self-promotion, or spam.