Biography
Julien Hauret is a French researcher in machine learning and speech processing, currently pursuing a Ph.D. at the Conservatoire National des Arts et Métiers (Cnam) in Paris. His doctoral research focuses on applying deep learning techniques to enhance speech signals captured via body-conduction microphones, aiming to improve communication in noisy environments.
In 2019, Hauret completed a four-month research internship at Columbia University's Heffner Biomedical Imaging Lab (HBIL) as part of his undergraduate studies at École normale supérieure Paris-Saclay. During this internship, he worked on using ML to learn a dictionary of textons on two large cohorts of CT imaging data.
Hauret has co-authored several publications, including the development of the Vibravox dataset, which comprises French speech recordings captured with body-conduction audio sensors. This dataset supports research in speech recognition, enhancement, and speaker verification.