Speech recognition using non-linear trajectories in a formant-based articulatory layer of a multiple-level segmental HMM.
Hongwei HuMartin J. RussellPublished in: INTERSPEECH (2008)
Keyphrases
- speech recognition
- speech signal
- hidden markov models
- vocal tract
- automatic speech recognition
- multi stream
- speech synthesis
- automatic speech recognition systems
- speaker identification
- keyword spotting
- speech recognizer
- noisy environments
- speech processing
- speaker independent
- pattern recognition
- language model
- acoustic features
- speech recognition systems
- conditional random fields
- levels of abstraction
- gesture recognition
- linear prediction
- neural network
- speaker adaptation
- lower level
- recognition engine
- trajectory data
- multiple targets
- non stationary
- spatio temporal
- image sequences