Research

Find my full publication list on

• ORCID
• G. Scholar

To learn more, visit the accompanying website to my doctoral thesis to learn more about my research conducted at IRCAM. I have worked on

• Glottal source detection and estimation
• Neural vocoders
• Developed a framework for disentanglement of voice attributes
• Estimation of voice level (vocal effort / intensity / musical dynamics in singing)

I’m currently a researcher at Supertone, where I maintain and develop our voice backbone, NANSY, used for text-to-speech and voice conversion. My work includes

• Self-supervised pitch estimation (in particular in the presence of sub-harmonics)
• Room acoustics estimation and modeling
• Estimation and transformation of other voice attributes