Deep Scattering Power Spectrum Features for Robust Speech Recognition.
Neethu M. JoyDino OglicZoran CvetkovicPeter BellSteve RenalsPublished in: INTERSPEECH (2020)
Keyphrases
- speech recognition
- power spectrum
- speech recognition systems
- noisy environments
- hidden markov models
- language model
- power spectral
- cepstral coefficients
- speech synthesis
- speech signal
- speech processing
- natural images
- automatic speech recognition
- pattern recognition
- neural network
- speaker identification
- multiresolution
- speech recognizer
- speech recognition technology
- speaker independent
- high order statistics
- density function
- frequency domain
- feature space
- mel frequency cepstral coefficients
- computational complexity
- data mining