Exploiting Hidden Representations from a DNN-based Speech Recogniser for Speech Intelligibility Prediction in Hearing-impaired Listeners.
Zehai TuNing MaJon BarkerPublished in: INTERSPEECH (2022)
Keyphrases
- hearing impaired
- speech recognition
- speech signal
- linear prediction
- prediction accuracy
- audio visual
- broadcast news
- speech synthesis
- text to speech
- data sets
- data mining
- speech recognizer
- speech processing
- speaker identification
- automatic speech recognition systems
- vocal tract
- noise reduction
- signal to noise ratio
- multi modal
- low level
- case study
- multimedia
- social networks
- neural network