Home // ACHI 2026, The Nineteenth International Conference on Advances in Computer-Human Interactions // View article

Event-Aware Audio Generation for LLM-Driven Storytelling in Extended Reality

Authors:
Mehmet Karaaslan
Meral Kuyucu
Bora Şenceylan
Gökhan İnce

Keywords: extended reality; generative audio; LLM-based systems; event-aware audio; immersive storytelling.

Abstract:
Sound is essential for immersion in Extended Reality (XR), yet audio design is often manual and disconnected from narrative context. This paper presents a Large Language Model (LLM) driven pipeline for event-aware audio generation in XR storytelling. The system extracts sound-inducing events from scene inputs, generates context-specific audio using a diffusion model, and selects high-quality samples through an LLM-based judging mechanism. Generated sounds are bound to narrative events using timing and repetition cues. We evaluated the system using a six scene story displayed in XR with 20 participants. Results show improved narrative-audio alignment and immersion when using Semantic Representation.

Pages: 120 to 125

Copyright: Copyright (c) IARIA, 2026

Publication date: May 24, 2026

Published in: conference

ISSN: 2308-4138

ISBN: 978-1-68558-383-5

Location: Venice, Italy

Dates: from May 24, 2026 to May 28, 2026