DIML: Deep Interpretable Metric Learning via Structural Matching.
Wenliang ZhaoYongming RaoJie ZhouJiwen LuPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2024)
Keyphrases
- metric learning
- structural matching
- distance metric
- distance metric learning
- pairwise
- semi supervised
- machine learning and pattern recognition
- learning tasks
- distance function
- multi task
- person re identification
- dimensionality reduction
- semi supervised clustering
- feature space
- semi supervised learning
- mahalanobis metric
- maximum variance unfolding
- labeled data
- nearest neighbor classification
- neural network
- learning problems
- unlabeled data
- active learning
- training set
- margin maximization
- feature extraction
- database systems