analiza aiqbench
This paper pairs a compact set of hand-engineered singing descriptors (prosody, spectral, phonetic, and vocal technique features) with SHAP and Grad-CAM to explain which cues drive emotion predictions. Its most useful finding for ML readers is that domain-specific features may reduce cross-singer variability and yield more interpretable models than end-to-end deep networks, though the evidence rests on a relatively small set of datasets and qualitative Grad-CAM analysis. Readers should watch whether the interpretability gains hold up on larger, more diverse corpora.
Interpretable Singing Emotion Recognition Using Acoustic and Vocal Technique Features
arXiv:2610.32629v1 Announce Type: cross Abstract: Singing emotion recognition relies predominantly on deep learning models that achieve strong performance but offer limited insight into the vocal characteristics driving emotion, of- ten overlooking the singing voice as a distinct signal. We propose an interpretable singing emotion recognition framework using a compact set of singing-specific descriptors across prosody, spec- tral features, phonetics, and vocal technique. To decode model decisions and link predictions to measurable acoustic phenomena, we employ SHAP and apply GRAD-CAM to qualitatively assess the learned representations of vocal techniques. Evaluations show that singing-specific features reduce cross-singer variability, while vocal technique descriptors provide interpretable insights into singing expression. Furthermore, out-of-domain evaluation on VocalSet reveals interesting associations between emotional expression and ground-truth technical execution. The analysis highlights vibrato and pharyngeal techniques as salient cues of emotional delivery, with SHAP indicating that vibrato is context-dependent on articulation rate. These findings emphasize the value of domain-specific features for interpretable singing emotion analysis.
Abstract:Singing emotion recognition relies predominantly on deep learning models that achieve strong performance but offer limited insight into the vocal characteristics driving emotion, of- ten overlooking the singing voice as a distinct signal. We propose an interpretable singing emotion recognition framework using a compact set of singing-specific descriptors across prosody, spec- tral features, phonetics, and vocal technique. To decode model decisions and link predictions to measurable acoustic phenomena, we employ SHAP and apply GRAD-CAM to qualitatively assess the learned representations of vocal techniques. Evaluations show that singing-specific features reduce cross-singer variability, while vocal technique descriptors provide interpretable insights into singing expression. Furthermore, out-of-domain evaluation on VocalSet reveals interesting associations between emotional expression and ground-truth technical execution. The analysis highlights vibrato and pharyngeal techniques as salient cues of emotional delivery, with SHAP indicating that vibrato is context-dependent on articulation rate. These findings emphasize the value of domain-specific features for interpretable singing emotion analysis.
Sound (cs.SD); Artificial Intelligence (cs.AI)
arXiv:2610.32629 [cs.SD]
(or arXiv:2610.32629v1 [cs.SD] for this version)
https://doi.org/10.48550/arXiv.2610.32629 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 248 cuvinte
Subiecte conexe
CARE: Constrained Attention Refinement for Fine-Grained Visual Classification via Teacher-Student Distillation
arXiv · 10 oct. 2026
Beyond Vision-Language Models: Visual-Semantic Affect Graphs for Emotion Recognition in Artworks
arXiv · 27 oct. 2026
Multi-Task Learning for Non-Canonical Phoneme Recognition via Articulatory Feature Decomposition
arXiv · 25 aug. 2026
Design of a Human-Assistance Robot System with Contextual Action Recognition
arXiv · 25 aug. 2026