Speaker and Direction Inferred Dual-Channel Speech Separation.
Chenxing LiJiaming XuNima MesgaraniBo XuPublished in: ICASSP (2021)
Keyphrases
- dual channel
- speech recognition
- audio visual
- speaker recognition
- automatic speech recognition
- sound source
- speaker verification
- speaker identification
- speech signal
- speaker diarization
- vocal tract
- automatic speech recognition systems
- prosodic features
- speaker dependent
- gaussian mixture model
- speech synthesis
- hidden markov models
- synthesized speech
- multi modal
- noisy environments
- pattern recognition
- text to speech
- automatic transcription
- recognition engine
- speech recognition systems
- acoustic features
- audio stream
- dialogue system
- speaker independent
- vector quantization
- spoken language
- endpoint detection
- neural network
- language model
- speech recognizer
- multi stream
- single channel
- non stationary
- broadcast news
- linear prediction