Squeezing Value of Cross-Domain Labels: A Decoupled Scoring Approach for Speaker Verification.
Lantian LiYang ZhangJiawen KangThomas Fang ZhengDong WangPublished in: ICASSP (2021)
Keyphrases
- cross domain
- speaker verification
- noisy environments
- text categorization
- transfer learning
- multi label
- audio visual
- knowledge transfer
- multilayer perceptron
- language identification
- emotion recognition
- training data
- e government
- multi modal
- pairwise
- human computer interaction
- knn
- data mining
- target domain
- using artificial neural networks
- training set