Revisiting Training Strategies and Generalization Performance in Deep Metric Learning

TL;DR

This paper systematically revisits deep metric learning, analyzing training strategies and their impact on generalization, and proposes a regularization method to improve model performance on benchmarks.

Contribution

It provides a consistent comparison of DML objectives, uncovers the link between embedding density and generalization, and introduces a regularization technique to enhance DML performance.

Findings

DML objectives saturate more than previously reported.

Embedding space density correlates with generalization performance.

A simple regularization improves ranking-based DML models on benchmarks.

Abstract

Deep Metric Learning (DML) is arguably one of the most influential lines of research for learning visual similarities with many proposed approaches every year. Although the field benefits from the rapid progress, the divergence in training protocols, architectures, and parameter choices make an unbiased comparison difficult. To provide a consistent reference point, we revisit the most widely used DML objective functions and conduct a study of the crucial parameter choices as well as the commonly neglected mini-batch sampling process. Under consistent comparison, DML objectives show much higher saturation than indicated by literature. Further based on our analysis, we uncover a correlation between the embedding space density and compression to the generalization performance of DML models. Exploiting these insights, we propose a simple, yet effective, training regularization to reliably boost the performance of ranking-based DML models on various standard benchmark datasets. Code and a publicly accessible WandB-repo are available at https://github.com/Confusezius/Revisiting_Deep_Metric_Learning_PyTorch.