Deep Ensemble Collaborative Learning by using Knowledge-transfer Graph for Fine-grained Object Classification

TL;DR

This paper introduces a novel ensemble learning method using a knowledge-transfer graph with attention maps to promote diversity among networks, significantly improving fine-grained object classification accuracy.

Contribution

It proposes an innovative ensemble approach with optimized knowledge transfer graphs and attention-based knowledge, enhancing diversity and accuracy in mutual learning ensembles.

Findings

Outperforms conventional ensemble methods on Stanford datasets.

Optimized knowledge-transfer graphs improve ensemble accuracy.

Attention-based knowledge transfer enhances diversity among networks.

Abstract

Mutual learning, in which multiple networks learn by sharing their knowledge, improves the performance of each network. However, the performance of ensembles of networks that have undergone mutual learning does not improve significantly from that of normal ensembles without mutual learning, even though the performance of each network has improved significantly. This may be due to the relationship between the knowledge in mutual learning and the individuality of the networks in the ensemble. In this study, we propose an ensemble method using knowledge transfer to improve the accuracy of ensembles by introducing a loss design that promotes diversity among networks in mutual learning. We use an attention map as knowledge, which represents the probability distribution and information in the middle layer of a network. There are many ways to combine networks and loss designs for knowledge transfer methods. Therefore, we use the automatic optimization of knowledge-transfer graphs to consider a variety of knowledge-transfer methods by graphically representing conventional mutual-learning and distillation methods and optimizing each element through hyperparameter search. The proposed method consists of a mechanism for constructing an ensemble in a knowledge-transfer graph, attention loss, and a loss design that promotes diversity among networks. We explore optimal ensemble learning by optimizing a knowledge-transfer graph to maximize ensemble accuracy. From exploration of graphs and evaluation experiments using the datasets of Stanford Dogs, Stanford Cars, and CUB-200-2011, we confirm that the proposed method is more accurate than a conventional ensemble method.