Lightweight Structured Multimodal Reasoning for Clinical Scene Understanding in Robotics

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jha, Saurav, Ehrlich, Stefan K.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912608072761344
author Jha, Saurav
Ehrlich, Stefan K.
author_facet Jha, Saurav
Ehrlich, Stefan K.
contents Healthcare robotics requires robust multimodal perception and reasoning to ensure safety in dynamic clinical environments. Current Vision-Language Models (VLMs) demonstrate strong general-purpose capabilities but remain limited in temporal reasoning, uncertainty estimation, and structured outputs needed for robotic planning. We present a lightweight agentic multimodal framework for video-based scene understanding. Combining the Qwen2.5-VL-3B-Instruct model with a SmolAgent-based orchestration layer, it supports chain-of-thought reasoning, speech-vision fusion, and dynamic tool invocation. The framework generates structured scene graphs and leverages a hybrid retrieval module for interpretable and adaptive reasoning. Evaluations on the Video-MME benchmark and a custom clinical dataset show competitive accuracy and improved robustness compared to state-of-the-art VLMs, demonstrating its potential for applications in robot-assisted surgery, patient monitoring, and decision support.
format Preprint
id arxiv_https___arxiv_org_abs_2509_22014
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Lightweight Structured Multimodal Reasoning for Clinical Scene Understanding in Robotics
Jha, Saurav
Ehrlich, Stefan K.
Computer Vision and Pattern Recognition
Artificial Intelligence
Human-Computer Interaction
Robotics
Healthcare robotics requires robust multimodal perception and reasoning to ensure safety in dynamic clinical environments. Current Vision-Language Models (VLMs) demonstrate strong general-purpose capabilities but remain limited in temporal reasoning, uncertainty estimation, and structured outputs needed for robotic planning. We present a lightweight agentic multimodal framework for video-based scene understanding. Combining the Qwen2.5-VL-3B-Instruct model with a SmolAgent-based orchestration layer, it supports chain-of-thought reasoning, speech-vision fusion, and dynamic tool invocation. The framework generates structured scene graphs and leverages a hybrid retrieval module for interpretable and adaptive reasoning. Evaluations on the Video-MME benchmark and a custom clinical dataset show competitive accuracy and improved robustness compared to state-of-the-art VLMs, demonstrating its potential for applications in robot-assisted surgery, patient monitoring, and decision support.
title Lightweight Structured Multimodal Reasoning for Clinical Scene Understanding in Robotics
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Human-Computer Interaction
Robotics
url https://arxiv.org/abs/2509.22014