DeepMind research scientist Dr. Kory Mathewson and Dr. Gauthier Gidel, professor at the Université de Montréal and Mila faculty member, discuss how new datasets like the ExVo challenge data are essential to progress in understanding vocal expression.Hume AI is proud to co-organize ICML's Expressive Vocalization (ExVo) Workshop and Competition. Bringing together leading emotion scientists and AI researchers, ExVo explores the machine learning problem of understanding and generating vocal bursts—a wide range of expressive nonverbal vocalizations such as laughs, sighs, grunts, cries, screams, oohs, ahhs, and huhs (to name but a few). The challenge is co-sponsored by Hume AI, Mila - Quebec Artificial Intelligence Institute, and the National Film Board of Canada.
ExVo participants will take on up to three ML tasks using a newly introduced, large-scale dataset with 50k recordings from 1.7k speakers. The dataset draws on innovations in emotion science to capture 10 dimensions of emotion reliably perceived in distinct vocal bursts: Awe, Excitement, Amusement, Awkwardness, Fear, Horror, Distress, Triumph, Sadness and Surprise. The tasks include Multi-Task, Generative, and Few-Shot Learning strategies. Given that the dataset is the first of its kind, we anticipate that the models that come out of this competition will be entirely new.
The ExVo Workshop will also be accepting contributions on other related topics:
-Detecting and Understanding Vocal Emotional Behavior
-Multi-Task Learning in Affective Computing
-Generating Nonverbal Vocalizations or Speech Prosody
-Personalized Machine Learning for Affective Computing
-Other topics related to Affective Verbal and Nonverbal Vocalization
For more information, rules, and deadlines visit: bit.ly/icmlexvo
About Hume AI: Hume AI is an AI research lab and technology company. Our mission is to ensure that future technology is built to serve our emotions, ensuring that as AI gets smarter it remains aligned with our goals.
Музыка
Фильмы
ТВ Шоу
Категории видео
