A second-order HMM for high performance word and phoneme-based continuous speech recognition.
Jean-François MariDominique FohrJean-Claude JunquaPublished in: ICASSP (1996)
Keyphrases
- continuous speech recognition
- hidden markov models
- speech recognition
- keyword spotting
- speech recognizer
- speech recognition systems
- handwritten words
- automatic speech recognition
- automatic speech recognition systems
- speech signal
- higher order
- speech synthesis
- prosodic features
- co occurrence
- phoneme recognition
- word recognition
- n gram
- speaker adaptation
- handwritten word recognition
- cost effective
- speech sounds
- handwriting recognition
- related words
- language model
- natural language text
- speaker independent
- multi stream
- visual speech
- multiword
- unknown words
- scientific computing
- feature selection
- high efficiency
- context dependent
- maximum likelihood
- pattern recognition
- keywords