clement-pages/pyannote-speech-separation-pipeline · Apply for community grant: Academic project (gpu)

Hello,

This space is dedicated to a demonstration of the use of a pipeline, which takes an audio conversation and outputs the corresponding diarization (who is speaking when), as well as the separate sources corresponding to each speaker in the conversation. These outputs can then be used by an Automatic Speech Recognition system, for example. This pipeline needs to run on a GPU to provide results within an acceptable amount of time, which is why I'm making this request for access to a Zero GPU.

Have a nice day!