Guided Deep Metric Learning

TL;DR

This paper introduces Guided Deep Metric Learning, a new architecture that enhances generalization in visual similarity learning by using a multi-branch model and knowledge distillation to better capture the data manifold, especially under distributional shifts.

Contribution

It proposes a novel architecture combining a multi-branch model and offline knowledge distillation to improve manifold generalization in deep metric learning.

Findings

Up to 40% improvement in Recall@1 on CIFAR10.

Better generalization under distributional shifts.

Enhanced data manifold representation.

Abstract

Deep Metric Learning (DML) methods have been proven relevant for visual similarity learning. However, they sometimes lack generalization properties because they are trained often using an inappropriate sample selection strategy or due to the difficulty of the dataset caused by a distributional shift in the data. These represent a significant drawback when attempting to learn the underlying data manifold. Therefore, there is a pressing need to develop better ways of obtaining generalization and representation of the underlying manifold. In this paper, we propose a novel approach to DML that we call Guided Deep Metric Learning, a novel architecture oriented to learning more compact clusters, improving generalization under distributional shifts in DML. This novel architecture consists of two independent models: A multi-branch master model, inspired from a Few-Shot Learning (FSL) perspective, generates a reduced hypothesis space based on prior knowledge from labeled data, which guides or regularizes the decision boundary of a student model during training under an offline knowledge distillation scheme. Experiments have shown that the proposed method is capable of a better manifold generalization and representation to up to 40% improvement (Recall@1, CIFAR10), using guidelines suggested by Musgrave et al. to perform a more fair and realistic comparison, which is currently absent in the literature