Acoustic characterization of speech rhythm: going beyond metrics with recurrent neural networks.
François DelocheLaurent Bonnasse-GahotJudit GervainPublished in: CoRR (2024)
Keyphrases
- recurrent neural networks
- speech sounds
- speech recognition systems
- audio signal
- acoustic features
- speech recognition
- emotional speech
- neural network
- complex valued
- echo state networks
- speech signal
- prosodic features
- recurrent networks
- feed forward
- formant frequencies
- speaker independent
- acoustic models
- speaker identification
- reservoir computing
- speech synthesis
- emotion recognition
- feedforward neural networks
- automatic speech recognition
- acoustic signal
- nonlinear dynamic systems
- text to speech
- speech recognizers
- spoken language
- vowel phonemes
- vocal tract
- audio visual
- artificial neural networks
- cascade correlation
- neural model
- visual speech
- speech recognizer
- spontaneous speech
- speaker recognition
- hidden markov models
- long term
- knowledge base