Discriminatively trained joint speaker and environment representations for adaptation of deep neural network acoustic models.
Maofan YinSunil SivadasKai YuBin MaPublished in: ICASSP (2016)
Keyphrases
- discriminatively trained
- discriminative training
- acoustic models
- neural network
- speech recognition
- automatic speech recognition
- object detection
- hidden markov models
- discriminative learning
- pattern recognition
- sequence classification
- maximum likelihood
- posterior probability
- generative model
- pairwise
- broadcast news