Safe Guard: an LLM-agent for Real-time Voice-based Hate Speech Detection in Social Virtual Reality

TL;DR

Safe Guard is a real-time voice-based hate speech detection system in social VR using LLMs and audio features, demonstrating improved accuracy and reduced false positives to promote safer virtual environments.

Contribution

The paper introduces a novel LLM-agent system for real-time hate speech detection in social VR, combining GPT and audio features, with evaluation showing enhanced performance over existing methods.

Findings

Effective real-time hate speech detection in social VR

Reduced false positive rate compared to existing approaches

Potential for safer virtual social environments

Abstract

In this paper, we present Safe Guard, an LLM-agent for the detection of hate speech in voice-based interactions in social VR (VRChat). Our system leverages Open AI GPT and audio feature extraction for real-time voice interactions. We contribute a system design and evaluation of the system that demonstrates the capability of our approach in detecting hate speech, and reducing false positives compared to currently available approaches. Our results indicate the potential of LLM-based agents in creating safer virtual environments and set the groundwork for further advancements in LLM-driven moderation approaches.