Semi-supervised Multichannel Speech Separation Based on a Phone- and Speaker-Aware Deep Generative Model of Speech Spectrograms.
Yicheng DuKouhei SekiguchiYoshiaki BandoAditya Arie NugrahaMathieu FontaineKazuyoshi YoshiiTatsuya KawaharaPublished in: EUSIPCO (2020)
Keyphrases
- generative model
- semi supervised
- speech recognition
- speech signal
- audio visual
- automatic speech recognition
- speaker recognition
- speaker identification
- speaker verification
- acoustic models
- vocal tract
- mixture model
- probabilistic model
- speech synthesis
- semi supervised learning
- linear prediction
- discriminative learning
- speaker diarization
- prior knowledge
- bayesian framework
- latent dirichlet allocation
- hidden markov models
- supervised learning
- text to speech
- multi view
- labeled data
- speaker dependent
- broadcast news
- spoken language
- noisy environments
- unlabeled data
- topic models
- pairwise
- prosodic features
- discriminative models
- emotion recognition
- hierarchical bayesian model
- speaker adaptation
- training data