Robust Acoustic Speech Feature Prediction From Noisy Mel-Frequency Cepstral Coefficients.
Ben MilnerJonathan DarchPublished in: IEEE Trans. Speech Audio Process. (2011)
Keyphrases
- mel frequency cepstral coefficients
- noisy environments
- acoustic features
- speech signal
- speaker identification
- feature set
- feature vectors
- speaker recognition
- speech recognition
- speaker verification
- cepstral coefficients
- linear predictive
- speech recognition systems
- gaussian mixture model
- linear prediction
- spectral features
- automatic speech recognition
- visual features
- audio features
- principal component analysis
- prediction accuracy
- prediction error
- noise reduction
- extracting features
- image features
- audio signal
- probabilistic neural network
- classification accuracy
- feature extraction
- speaker diarization
- information retrieval
- language model
- broadcast news
- hidden markov models
- vector quantization
- audio visual
- edge detection
- image processing
- feature selection