Speaker-Aware Training of Attention-Based End-to-End Speech Recognition Using Neural Speaker Embeddings.
Aku RouheTuomas KasevaMikko KurimoPublished in: ICASSP (2020)
Keyphrases
- end to end
- speech recognition
- speaker recognition
- audio visual
- speaker verification
- automatic speech recognition
- acoustic models
- speaker identification
- speaker diarization
- speaker dependent
- prosodic features
- speech signal
- vocal tract
- speech synthesis
- admission control
- congestion control
- broadcast news
- multipath
- neural network
- wireless ad hoc networks
- automatic transcription
- network architecture
- synthesized speech
- automatic speech recognition systems
- high bandwidth
- speech recognizer
- acoustic features
- content delivery
- noisy environments
- hidden markov models
- speech sounds
- internet protocol
- visual speech
- scalable video
- speaker adaptation
- ad hoc networks