No Pairs Left Behind: Improving Metric Learning with Regularized Triplet Objective.
A. Ali HeydariNaghmeh RezaeiDaniel J. McDuffJavier L. PrietoPublished in: CoRR (2022)
Keyphrases
- metric learning
- pairwise
- distance metric
- distance metric learning
- pairwise similarities
- nearest neighbor classification
- semi supervised
- machine learning and pattern recognition
- multi task
- learning tasks
- dimensionality reduction
- mahalanobis metric
- distance function
- semi supervised learning
- multi class
- mahalanobis distance
- maximum variance unfolding
- feature space
- similarity measure
- domain knowledge
- query processing
- machine learning