Unsupervised Sound Separation Using Mixture Invariant Training.
Scott WisdomEfthymios TzinisHakan ErdoganRon J. WeissKevin W. WilsonJohn R. HersheyPublished in: NeurIPS (2020)
Keyphrases
- supervised learning
- supervised training
- training set
- unsupervised learning
- machine learning
- blind separation
- multiscale
- deep architectures
- restricted boltzmann machine
- sound source
- training algorithm
- mixture model
- data driven
- semi supervised
- affine transformation
- training phase
- test set
- training samples
- finite mixture model