Multitask Detection of Speaker Changes, Overlapping Speech and Voice Activity Using Wav2vec 2.0.
Marie KunesováZbynek ZajícPublished in: ICASSP (2023)
Keyphrases
- multi task
- voice activity detection
- speech recognition
- speaker verification
- prosodic features
- noisy environments
- text to speech
- speaker recognition
- audio visual
- speech sounds
- automatic speech recognition
- multitask learning
- synthesized speech
- speaker identification
- emotion recognition
- multi task learning
- speech signal
- speech synthesis
- learning tasks
- speaker dependent
- mel frequency cepstral coefficients
- multiple tasks
- speech quality
- hidden markov models
- feature selection
- acoustic features
- machine learning
- transfer learning
- learning experience
- probabilistic model