Probabilistic Permutation Invariant Training for Speech Separation.
Midia YousefiSoheil KhorramJohn H. L. HansenPublished in: INTERSPEECH (2019)
Keyphrases
- hearing impaired
- speech recognition
- training process
- training set
- speech signal
- probabilistic model
- generative model
- bayesian networks
- supervised learning
- training examples
- natural language
- probabilistic interpretation
- uncertain data
- automatic speech recognition
- moment invariants
- training phase
- conditional probabilities
- noisy environments
- machine learning
- affine invariant
- posterior probability
- neural network
- multi modal
- hidden markov models
- feature space
- support vector