Network of Experts for Large-Scale Image Categorization

TL;DR

This paper introduces a tree-structured network architecture called 'network of experts' for large-scale image classification, which improves accuracy by dividing the classification task into specialized branches while sharing common features.

Contribution

The paper proposes a novel end-to-end trainable tree-structured CNN architecture that automatically learns category partitions and improves classification accuracy with minimal additional computational cost.

Findings

Significant accuracy improvements on CIFAR100 and ImageNet datasets.

The method enhances existing CNNs with little increase in training time.

Achieved the best results on CIFAR100 among comparable methods.

Abstract

We present a tree-structured network architecture for large scale image classification. The trunk of the network contains convolutional layers optimized over all classes. At a given depth, the trunk splits into separate branches, each dedicated to discriminate a different subset of classes. Each branch acts as an expert classifying a set of categories that are difficult to tell apart, while the trunk provides common knowledge to all experts in the form of shared features. The training of our "network of experts" is completely end-to-end: the partition of categories into disjoint subsets is learned simultaneously with the parameters of the network trunk and the experts are trained jointly by minimizing a single learning objective over all classes. The proposed structure can be built from any existing convolutional neural network (CNN). We demonstrate its generality by adapting 4 popular CNNs for image categorization into the form of networks of experts. Our experiments on CIFAR100 and ImageNet show that in every case our method yields a substantial improvement in accuracy over the base CNN, and gives the best result achieved so far on CIFAR100. Finally, the improvement in accuracy comes at little additional cost: compared to the base network, the training time is only moderately increased and the number of parameters is comparable or in some cases even lower.