Arabic Dysarthric Speech Recognition Using Adversarial and Signal-Based Augmentation.
Massa BaaliIbrahim AlmakkyShady ShehataFakhri KarrayPublished in: CoRR (2023)
Keyphrases
- speech recognition
- speech quality
- speech corpus
- signal processing
- speech signal
- noisy environments
- endpoint detection
- high frequency
- non stationary
- arabic language
- automatic speech recognition
- spoken language
- fundamental frequency
- acoustic signal
- impulse response
- broadcast news
- hearing aids
- hearing impaired
- heart rate
- compressive sensing
- audio visual
- frequency domain
- multi agent
- speech synthesis
- spoken document retrieval
- signal detection
- recognition engine
- dialogue system
- arabic text
- speech segments
- isolated word
- hidden markov models
- multiresolution
- feature vectors