Canonicalization of Feature Parameters for Robust Speech Recognition Based on Distinctive Phonetic Feature (DPF) Vectors.
Mohammad Nurul HudaMuhammad GhulamTakashi FukudaKouichi KatsuradaTsuneo NittaPublished in: IEICE Trans. Inf. Syst. (2008)
Keyphrases
- speech recognition
- feature vectors
- hidden markov models
- speech understanding
- speech synthesis
- pattern recognition
- noisy environments
- automatic speech recognition
- speech processing
- speech recognition technology
- language model
- cepstral coefficients
- speech recognizer
- speaker diarization
- speech recognition systems
- neural network
- speech recognizers
- speaker identification
- speech signal
- mobile devices