Acoustic-to-Articulatory Speech Inversion Features for Mispronunciation Detection of /ɹ/ in Child Speech Sound Disorders.
Nina R. BenwayYashish M. SiriwardenaJonathan L. PrestonElaine HitchcockTara McAllister ByunCarol Y. Espy-WilsonPublished in: INTERSPEECH (2023)
Keyphrases
- acoustic features
- speech recognition
- speech recognition systems
- speech signal
- mel frequency cepstral coefficients
- emotional speech
- audio features
- automatic speech recognition
- speech synthesis
- vocal tract
- speech segments
- speech sounds
- emotion recognition
- hidden markov models
- false positives
- audio signal
- sound source
- formant frequencies
- visual speech
- automatic speech recognition systems
- speaker verification
- spectral features
- speech recognizer
- music information retrieval
- noisy environments
- audio visual
- environmental sounds
- visual features
- multi stream
- low level
- lexical features
- recognition engine
- text to speech
- image features
- speaker recognition
- broadcast news
- spoken language
- speaker independent
- noisy speech
- emotion classification
- feature space
- pattern recognition
- image retrieval
- speech recognizers
- anomaly detection
- audio stream
- detection algorithm
- detection rate
- visual speech recognition
- prosodic features
- extracted features
- speech enhancement
- dialogue system
- source localization