Hybrid generative-discriminative training of Gaussian mixture models.
Wolfgang RothRobert PeharzSebastian TschiatschekFranz PernkopfPublished in: Pattern Recognit. Lett. (2018)
Keyphrases
- gaussian mixture model
- hybrid generative discriminative
- mixture model
- speaker recognition
- em algorithm
- maximum likelihood
- feature vectors
- expectation maximization
- discriminative models
- training set
- generative model
- feature space
- information retrieval
- supervised learning
- covariance matrices
- unsupervised learning
- semi supervised learning
- speaker identification
- object recognition
- image processing