StutterNet: Stuttering Disfluencies Detection in Synthetic Speech Signals via Mel Frequency Cepstral Coefficients Features Using Deep Learning.
Muhammad AbubakarMuhammad MujahidKhadija KanwalSajid IqbalMuhammad Nabeel AsgharAbdullah A. AlaulamiePublished in: IEEE Access (2024)
Keyphrases
- speech signal
- mel frequency cepstral coefficients
- deep learning
- automatic speech recognition
- speaker identification
- speech recognition
- noisy environments
- speaker recognition
- cepstral coefficients
- acoustic features
- feature vectors
- hidden markov models
- linear prediction
- low level
- feature set
- object detection
- blind source separation
- image features
- non stationary
- feature extraction
- extracting features
- gaussian mixture model
- spectral analysis
- unsupervised learning
- classification accuracy
- multiscale
- image processing
- noisy images
- feature selection
- principal component analysis
- spectral features
- probabilistic model