Dysarthric Speech Recognition Error Correction Using Weighted Finite State Transducers Based on Context-Dependent Pronunciation Variation.
Woo Kyeong SeongJi Hun ParkHong Kook KimPublished in: ICCHP (2) (2012)
Keyphrases
- speech recognition
- error correction
- context dependent
- finite state transducers
- hidden markov models
- language model
- language modeling
- pattern recognition
- natural language
- error detection
- speech signal
- automatic speech recognition
- finite state
- speech synthesis
- n gram
- noisy environments
- speech recognizer
- low level
- speech recognition systems
- speaker independent
- machine translation
- neural network
- acoustic models
- image processing
- computer vision
- natural language processing
- mobile devices
- high level
- feature selection
- information retrieval