An investigation on the degradation of different features extracted from the compressed American English speech using narrowband and wideband codecs.
M. S. Arun SankarP. S. SathideviPublished in: Int. J. Speech Technol. (2018)
Keyphrases
- speech signal
- extracted features
- automatically extracted
- speech recognition
- feature extraction
- mel frequency cepstral coefficients
- feature set
- natural language
- extracting features
- low level
- spoken language
- speaker identification
- textural features
- text to speech
- spectral features
- feature vectors
- non stationary
- signal to noise ratio
- co occurrence
- information extraction
- spectral analysis
- hidden markov models
- lexical features
- speech quality