GMM-derived features for effective unsupervised adaptation of deep neural network acoustic models.

Natalia A. Tomashenko Yuri Y. Khokhlov

Published in: INTERSPEECH (2015)

Keyphrases

neural network
feature vectors
feature set
image features
feature space
feature extraction
pattern recognition
supervised learning
gaussian mixture model
noise reduction
video sequences
hidden markov models
unsupervised learning
extracting features
probabilistic neural network
speaker recognition