L2 proficiency assessment using self-supervised speech representations.
Stefano BannòKate M. KnillMarco MatassoniVyas RainaMark J. F. GalesPublished in: CoRR (2022)
Keyphrases
- speech recognition
- speech signal
- higher level
- text to speech
- neural network
- multi modal
- endpoint detection
- recognition engine
- spoken language
- multiple representations
- assessment process
- speech synthesis
- automatic speech recognition systems
- database
- spontaneous speech
- english text
- speaker recognition
- foreign language
- noisy environments
- risk assessment
- data sets