Speaker Diarization Using Deep Recurrent Convolutional Neural Networks for Speaker Embeddings.

Pawel Cyrta Tomasz Trzcinski Wojciech Stokowiec

Published in: ISAT (1) (2017)

Keyphrases

speaker diarization
convolutional neural networks
convolutional network
speech recognition
broadcast news
bayesian information criterion
audio stream
neural network
dimensionality reduction
decision trees
high dimensional data
gaussian mixture model