Power Spectrum Difference Teager Energy Features for Speech Recognition in Noisy Environment.
Navnath S. NeheRaghunath S. HolambePublished in: ICIIS (2008)
Keyphrases
- speech recognition
- noisy environments
- power spectrum
- speech signal
- speech recognition systems
- automatic speech recognition
- speaker identification
- speaker verification
- language model
- speech enhancement
- pattern recognition
- hidden markov models
- background noise
- cepstral coefficients
- power spectral
- speech synthesis
- noise reduction
- frequency domain
- mel frequency cepstral coefficients
- speech recognizer
- voice activity detection
- natural images
- audio visual speech recognition
- noisy images
- noisy speech
- image processing
- feature extraction
- visual speech
- speaker diarization
- similarity measure
- multiscale
- extracting features
- feature selection