Evaluating Prompt Robustness in Text-to-Audio Systems for Adaptive Virtual Agents and Game Soundtracks
arXiv:2603.13824 · doi:10.1145/3806774.3832791
Abstract
Recent text-to-audio models enable adaptive game soundtracks, but small prompt changes can cause abrupt musical shifts. We evaluate MusicGen-small, MusicGen-large, and Stable Audio 2.5 under Minimal Lexical Substitution, Intensity Shifts, and Structural Rephrasing using log-Mel distance, MFCC/chroma-DTW, and CLAP similarity. Stable Audio 2.5 achieves the lowest pooled acoustic distances and the highest audio-audio CLAP similarity under structural rephrasing, while MusicGen-large has the highest audio-audio CLAP similarity under lexical substitutions and intensity shifts. Stable Audio 2.5 also shows the greatest between-seed variation in prompt-to-audio alignment, demonstrating the importance of multi-seed robustness evaluation for adaptive game audio.
arXiv comment4 pages, 2 figures. Poster paper published in the Proceedings of the ACM International Conference on Intelligent Virtual Agents (IVA 2026)