Metric Learning with Self-Adjusting Memory for Explaining Feature Drift.
Johannes KummertAlexander SchulzBarbara HammerPublished in: SN Comput. Sci. (2023)
Keyphrases
- metric learning
- multiple features
- distance metric
- machine learning and pattern recognition
- distance metric learning
- dimensionality reduction
- person re identification
- nearest neighbor classification
- pairwise
- mahalanobis metric
- distance function
- multi task
- learning tasks
- semi supervised clustering
- semi supervised
- image features
- feature space
- euclidean distance
- semi supervised learning
- feature vectors
- data sets
- background knowledge
- preprocessing step
- semi definite programming
- maximum variance unfolding
- similarity measure