Leveraging Pretrained ASR Encoders for Effective and Efficient End-to-End Speech Intent Classification and Slot Filling.
He HuangJagadeesh BalamBoris GinsburgPublished in: INTERSPEECH (2023)
Keyphrases
- end to end
- automatic speech recognition
- speech recognition
- admission control
- wireless ad hoc networks
- text localization and recognition
- speech signal
- high bandwidth
- ad hoc networks
- highly efficient
- noisy environments
- pattern recognition
- multimedia
- multipath
- content delivery
- feature vectors
- application layer
- congestion control
- spontaneous speech