Speech recognition in a noisy environment using a speech signal estimation method based on the Kalman filter.
Masakiyo FujimotoYasuo ArikiPublished in: Systems and Computers in Japan (2004)
Keyphrases
- speech recognition
- speech signal
- noisy environments
- kalman filter
- linear prediction
- automatic speech recognition
- noisy speech
- audio visual speech recognition
- spectral subtraction
- speaker recognition
- additive noise
- background noise
- hidden markov models
- speaker verification
- spectral analysis
- speaker identification
- language model
- hybrid method
- noise reduction
- non stationary
- wiener filter
- em algorithm
- vocal tract
- probabilistic model