Far-Field End-to-End Text-Dependent Speaker Verification Based on Mixed Training Data with Transfer Learning and Enrollment Data Augmentation.
Xiaoyi QinDanwei CaiMing LiPublished in: INTERSPEECH (2019)
Keyphrases
- end to end
- training data
- data sets
- transfer learning
- labeled data
- high dimensional data
- learning algorithm
- information retrieval
- test data
- unlabeled data
- knowledge discovery
- supervised learning
- neural network
- target domain
- semi supervised learning
- semi supervised
- data sources
- data analysis
- reinforcement learning
- image data
- data points
- active learning
- training samples
- training set
- feature space
- machine learning
- speaker verification