Vision-Language System using Open-Source LLMs for Gestures in Medical Interpreter Robots

TL;DR

This paper introduces a privacy-preserving vision-language system for medical robots that detects speech acts and generates gestures using open-source models, improving efficiency and human-likeness in healthcare communication.

Contribution

It presents a novel framework combining open-source LLMs and vision models for gesture generation in medical robots, along with a new annotated dataset for clinical conversations.

Findings

Identification module achieved 0.90 accuracy

System outperforms baseline in human-likeness

Maintains comparable appropriateness in gesture generation

Abstract

Effective communication is vital in healthcare, especially across language barriers, where non-verbal cues and gestures are critical. This paper presents a privacy-preserving vision-language framework for medical interpreter robots that detects specific speech acts (consent and instruction) and generates corresponding robotic gestures. Built on locally deployed open-source models, the system utilizes a Large Language Model (LLM) with few-shot prompting for intent detection. We also introduce a novel dataset of clinical conversations annotated for speech acts and paired with gesture clips. Our identification module achieved 0.90 accuracy, 0.93 weighted precision, and a 0.91 weighted F1-Score. Our approach significantly improves computational efficiency and, in user studies, outperforms the speech-gesture generation baseline in human-likeness while maintaining comparable appropriateness.