Abstract:Recently, substantial research efforts in Deep Metric Learning (DML) focused on designing complex pairwise-distance losses, which require convoluted schemes to ease optimization, such as sample mining or pair weighting. The standard cross-entropy loss for classification has been largely overlooked in DML. On the surface, the cross-entropy may seem unrelated and irrelevant to metric learning as it does not explicitly involve pairwise distances. However, we provide a theoretical analysis that links the cross-entropy to several well-known and recent pairwise losses. Our connections are drawn from two different perspectives: one based on an explicit optimization insight; the other on discriminative and generative views of the mutual information between the labels and the learned features. First, we explicitly demonstrate that the cross-entropy is an upper bound on a new pairwise loss, which has a structure similar to various pairwise losses: it minimizes intra-class distances while maximizing inter-class distances. As a result, minimizing the cross-entropy can be seen as an approximate bound-optimization (or Majorize-Minimize) algorithm for minimizing this pairwise loss. Second, we show that, more generally, minimizing the cross-entropy is actually equivalent to maximizing the mutual information, to which we connect several well-known pairwise losses. Furthermore, we show that various standard pairwise losses can be explicitly related to one another via bound relationships. Our findings indicate that the cross-entropy represents a proxy for maximizing the mutual information – as pairwise losses do – without the need for convoluted sample-mining heuristics. Our experiments (Code available at: https://github.com/jeromerony/dml_cross_entropy) over four standard DML benchmarks strongly support our findings. We obtain state-of-the-art results, outperforming recent and complex DML methods.

Decision Tree Pairwise Metric Learning Against Adversarial Attacks

Deep Adversarial Metric Learning

Metric Learning with Adaptive Density Discrimination

Distance-Dependent Metric Learning.

Improving Adversarial Robustness to Sensitivity and Invariance Attacks with Deep Metric Learning

Tree-Based Metric Learning For Distance Computation In Data Mining

Towards Adversarial Robustness for Multi-Mode Data through Metric Learning

A Distributed Approach Toward Discriminative Distance Metric Learning

A Unifying Mutual Information View of Metric Learning: Cross-Entropy vs. Pairwise Losses

Provably Robust Metric Learning

From Point to Set: Extend the Learning of Distance Metrics

Informative pairs mining based adaptive metric learning for adversarial domain adaptation

Toward Certified Robustness of Distance Metric Learning

metricDTW: local distance metric learning in Dynamic Time Warping

Deep Multiple Metric Learning for Time Series Classification

Signal-To-Noise Ratio: A Robust Distance Metric For Deep Metric Learning

Adversarial Metric Attack and Defense for Person Re-Identification

Adaptive neighborhood Metric learning

Decomposition-based transfer distance metric learning for image classification.

Deep Metric Learning with Tuplet Margin Loss.