Diversified Mutual Learning for Deep Metric Learning

TL;DR

This paper introduces Diversified Mutual Metric Learning, a novel approach that enhances deep metric learning models through diversified mutual knowledge transfer, significantly improving performance on standard datasets.

Contribution

It proposes a diversified mutual learning framework leveraging model, temporal, and view diversities to improve deep metric learning, especially in low-data transfer scenarios.

Findings

Achieves state-of-the-art Recall@1 on CUB-200-2011 and CARS-196 datasets.

Significantly improves individual and ensemble model performances.

Effective in inductive transfer learning with limited data.

Abstract

Mutual learning is an ensemble training strategy to improve generalization by transferring individual knowledge to each other while simultaneously training multiple models. In this work, we propose an effective mutual learning method for deep metric learning, called Diversified Mutual Metric Learning, which enhances embedding models with diversified mutual learning. We transfer relational knowledge for deep metric learning by leveraging three kinds of diversities in mutual learning: (1) model diversity from different initializations of models, (2) temporal diversity from different frequencies of parameter update, and (3) view diversity from different augmentations of inputs. Our method is particularly adequate for inductive transfer learning at the lack of large-scale data, where the embedding model is initialized with a pretrained model and then fine-tuned on a target dataset. Extensive experiments show that our method significantly improves individual models as well as their ensemble. Finally, the proposed method with a conventional triplet loss achieves the state-of-the-art performance of Recall@1 on standard datasets: 69.9 on CUB-200-2011 and 89.1 on CARS-196.