"Deep Learning Applied to SATB Choirs"


We present some recent research on the application of deep learning based methodologies to the field of Soprano, Alto, Tenor and Bass (SATB) choirs. Concretely, we present a dataset, an application of source separation to the data as well as multi-pitch estimation, analysis and synthesis of Unison form A Capella input and A Capella from Unison input.


I am Pritish Chandna, a PhD student in the fourth year at the Universitat Pompeu Fabra, under the tutelage of Emilia Gomez. My PhD focuses on the application of Deep Learning to the Singing Voice.

Singing_Voice_Presentation-1.pptx 12.7 MB, downloaded 56 times Polina Proutskova, 2020-09-22 11:36 AM