Multi-Task Learning of Speech Recognition and Speech Synthesis Parameters for Ultrasound-based Silent Speech Interfaces.
László TóthGábor GosztolyaTamás GrószAlexandra MarkóTamás Gábor CsapóPublished in: INTERSPEECH (2018)
Keyphrases
- speech synthesis
- speech recognition
- multi task learning
- multi task
- vocal tract
- speech signal
- automatic speech recognition
- speech recognizer
- learning tasks
- hidden markov models
- language model
- pattern recognition
- text to speech
- speech processing
- gaussian processes
- noisy environments
- neural network
- learning problems
- transfer learning
- speaker independent
- speech recognition systems
- high order
- maximum likelihood
- speech recognition technology
- isolated word
- theoretical analysis
- machine learning
- e learning
- learning algorithm
- data mining