Speech Enhancement Autoencoder with Hierarchical Latent Structure.
Koen OostermeijerJun DuQing WangChin-Hui LeePublished in: ICASSP (2021)
Keyphrases
- latent structure
- speech enhancement
- noise reduction
- multidimensional scaling
- noisy environments
- discriminative learning
- topic modeling
- latent variables
- negative matrix factorization
- single channel
- signal to noise ratio
- speech signal
- document classification
- generative model
- latent variable models
- maximum margin
- low dimensional
- bayesian networks
- vocal tract
- multiscale
- sparse representation
- nonnegative matrix factorization
- feature extraction
- multi channel