videos in AudioSet
A list of about a million ten-second YouTube clips, each labelled with the sounds in it, that Google published in 2017. Labs use it to train and test general audio and music models.
Including Demi-Finale III CMIM chant/voice 2015 : Hyesang Park, Demi-Finale I CMIM chant/voice 2015 : Claire de Sévigné