Joint tracking of clean speech and noise using HMMs and particle filters for robust speech recognition.
Aleem MushtaqChin-Hui LeePublished in: ACSCC (2012)
Keyphrases
- speech recognition
- particle filter
- noisy speech
- noisy environments
- robust tracking
- hidden markov models
- tracking framework
- particle filtering
- object tracking algorithm
- speech signal
- visual tracking
- object tracking
- spectral subtraction
- multiple object tracking
- automatic speech recognition
- speech recognizer
- speech synthesis
- speech enhancement
- background noise
- motion model
- state space
- speaker identification
- abrupt motion
- kalman filter
- speech processing
- language model
- partial occlusion
- appearance model
- mean shift
- pattern recognition
- sequential monte carlo
- signal to noise ratio
- speech recognition systems
- handwriting recognition
- speaker independent
- image sequences
- speech recognition errors
- speech recognition technology
- speech recognizers
- audio visual speech recognition
- speaker recognition
- articulated body
- neural network
- speaker dependent
- feature space
- face recognition