Computational Analysis and Simulation of the Human Voice (Dagstuhl Seminar 24242)

Abstract

This report documents the program and the outcomes of Dagstuhl Seminar 24242 "Computational Analysis and Simulation of the Human Voice", which was held from the 9th to the 14th of June, 2024. The seminar addressed key issues for a better understanding of the human voice by focusing on four main areas: voice analysis, visualisation techniques, simulation methods, and data analysis with machine learning. There has been enormous progress in recent years in all these fields. The seminar brought together a number of experts from fields as diverse as computer science, logopedics and phoniatrics, clinicians, acoustics and audio engineering, electronics, musicology, speech and hearing sciences, physics and mathematics. The schedule was quite flexible, including inspirational talks in the main areas, interactive working groups, sharing of conclusions and discussions, presentation of successes and failures to learn from, and a large number of free talks that emerged throughout the days. The variety of topics and participants created a highly enriching environment from which novel proposals for future research and collaboration emerged, as well as the collective writing of a paper on the state of the art and future perspectives in human voice research.