Mahalanobis Metric Scoring Learned from Weighted Pairwise Constraints in I-Vector Speaker Recognition System.
Zhenchun LeiYanhong WanJian LuoYingen YangPublished in: INTERSPEECH (2016)
Keyphrases
- pairwise constraints
- mahalanobis metric
- metric learning
- distance metric
- semi supervised
- pairwise
- spectral clustering
- feature extraction
- multi task
- data representation
- loss function
- distance function
- semi supervised learning
- learning tasks
- dimensionality reduction
- unlabeled data
- labeled data
- feature vectors
- decision trees
- covariance matrix
- euclidean distance
- document clustering
- distance measure
- decision boundary
- feature space