Domain adaptation using maximum likelihood linear transformation for PLDA-based speaker verification.
Qiongqiong WangHitoshi YamamotoTakafumi KoshinakaPublished in: ICASSP (2016)
Keyphrases
- linear transformation
- domain adaptation
- speaker verification
- maximum likelihood
- noisy environments
- cross domain
- multiple sources
- labeled data
- distance metric
- semi supervised
- expectation maximization
- audio visual
- semi supervised learning
- linear model
- transfer learning
- target domain
- unlabeled data
- learning tasks
- metric learning
- data sets
- speech recognition
- unsupervised learning
- text mining
- support vector machine
- active learning