Improved Silence-Unvoiced-Voiced (SUV) Segmentation for Dysarthric Speech Signals using Linear Prediction Error Variance.
Tolulope Bamidele IjitonaHong YueJohn J. SoraghanAnja LowitPublished in: ICCCS (2020)
Keyphrases
- speech signal
- prediction error
- linear prediction
- speech recognition
- automatic speech recognition
- speaker identification
- vocal tract
- motion compensated
- spectral analysis
- non stationary
- background noise
- noisy environments
- hidden markov models
- speech quality
- multiscale
- bit rate
- blind source separation
- automatic speech recognition systems
- inter frame
- motion vectors
- image segmentation
- multimedia
- sound source
- noisy images
- probabilistic model
- optical flow
- neural network
- reversible watermarking