Characterizing Generalization under Out-Of-Distribution Shifts in Deep Metric Learning

In this work, we systematically construct train-test splits of increasing difficulty and present the ooDML benchmark to characterize generalization under o ut-o f-distribution shifts in DML .