Characterizing Generalization under Out-Of-Distribution Shifts in Deep Metric Learning.
Timo MilbichKarsten RothSamarth SinhaLudwig SchmidtMarzyeh GhassemiBjörn OmmerPublished in: NeurIPS (2021)
Keyphrases
- metric learning
- distance metric
- distance metric learning
- pairwise
- dimensionality reduction
- machine learning and pattern recognition
- nearest neighbor classification
- multi task
- semi supervised
- learning tasks
- semi supervised clustering
- feature space
- distance function
- person re identification
- mahalanobis metric
- data sets
- maximum variance unfolding
- euclidean distance
- mahalanobis distance
- semi definite programming
- learning problems
- semi supervised learning
- high dimensional data
- image classification
- image features
- domain knowledge
- reinforcement learning
- knowledge base
- machine learning