Speaker diarization is the problem of identifying different speakers from a conversation and convert the same to speech-to-text format. This has been achieved here by affinity clustering method of sklearn.clusters

The dependency modules

1. Librosa (Sound processing and DSP)
2. Numpy (Matrix calculations)
3. Sklearn (ML Algorithms)
4. Scipy (Statistical calculations)

Python Version - 3.5 Audio format - wav