Online Streaming End-to-End Neural Diarization Handling Overlapping Speech and Flexible Numbers of Speakers.
Yawen XueShota HoriguchiYusuke FujitaYuki TakashimaShinji WatanabeLeibny Paola García-PereraKenji NagamatsuPublished in: Interspeech (2021)
Keyphrases
- end to end
- speech recognition
- scalable video
- speaker identification
- rate adaptation
- speaker dependent
- speaker diarization
- congestion control
- admission control
- multipath
- speech signal
- real time
- wireless ad hoc networks
- high bandwidth
- network architecture
- neural network
- automatic speech recognition
- ad hoc networks
- rate allocation
- content delivery
- stream processing
- packet losses
- internet protocol