Supervised vs. unsupervised learning of spectro temporal speech features.

Martin Heckmann

Published in: SAPA@INTERSPEECH (2010)

Keyphrases

unsupervised learning
supervised learning
feature set
learning algorithm
supervised training
feature extraction

spectral features
temporal information
spatio temporal
classification accuracy
co occurrence
low level

spatial features
temporal structure
feature space
hidden markov models
speech recognition

audio visual
unsupervised feature selection
speech synthesis
object recognition
lexical features