Saved in:
Bibliographic Details
Main Authors: Koh, Junyoung, Kim, Soo Yong, Choi, Gyu Hyeong, Choi, Yongwon
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.20891
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • We present AIBA (Attention-In-Band Alignment), a lightweight, training-free pipeline to quantify where text-to-audio diffusion models attend on the time-frequency (T-F) plane. AIBA (i) hooks cross-attention at inference to record attention probabilities without modifying weights; (ii) projects them to fixed-size mel grids that are directly comparable to audio energy; and (iii) scores agreement with instrument-band ground truth via interpretable metrics (T-F IoU/AP, frequency-profile correlation, and a pointing game). On Slakh2100 with an AudioLDM2 backbone, AIBA reveals consistent instrument-dependent trends (e.g., bass favoring low bands) and achieves high precision with moderate recall.