Low-dimensional space transforms of posteriors in speech recognition.
Jan ZelinkaJan TrmalLudek MüllerPublished in: INTERSPEECH (2010)
Keyphrases
- speech recognition
- low dimensional
- high dimensional
- speech synthesis
- hidden markov models
- speech processing
- dimensionality reduction
- pattern recognition
- language model
- automatic speech recognition
- speech recognition technology
- high dimensional data
- speech recognizer
- data points
- feature space
- speech recognition systems
- speech understanding
- speech recognizers
- speaker identification
- noisy environments
- principal component analysis
- speech signal
- keyword spotting
- speaker recognition
- speaker dependent
- gaussian mixture model
- non stationary
- machine learning