LAMAR-2 at MedGenVidQA 2026: Visual Answer Localization in Medical Videos via Multimodal LLM and Context-Augmented Prompting
Year: 2026
Type: conference-paper
Abstract: This paper presents an approach to localizing visual answers within continuous medical videos using a multi-step multimodal generation pipeline with the MedGenVidQA dataset.We frame visual answer localization as a multimodal fusion problem, integrating raw video, timestamped ASR transcripts, and VLMgenerated scene descriptions into structured contextual blocks, enabling the model to crossreference spoken commentary against observable physical events.We show that targeted guidance, which forces the model to treat audio transcripts as supplementary hints with observable visual movements, significantly outperforms baseline approaches.It achieves stateof-the-art performance on the test leaderboard, yielding an mIoU of 79.55, alongside [email protected],[email protected], and [email protected] scores of 93.75, 90.00, and 77.50, respectively.Our findings highlight the effectiveness of combining multimodal context fusion with targeted guidance to overcome text bias, establishing a promising approach for achieving the micro-level precision required in the medical domain.We release our code on GitHub at github.com/biodatlab/medgenvidqa- lamar.
Institutions: Mahidol University, Ramathibodi Hospital
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W7166873349