Automatically designing CNN architectures using genetic algorithm for image classification

TL;DR

This paper introduces an automatic CNN architecture design method using genetic algorithms, enabling users without domain expertise to generate effective models for image classification, outperforming existing automatic methods.

Contribution

The paper presents a novel genetic algorithm-based approach for automatic CNN architecture design that requires no domain knowledge from users.

Findings

Outperforms existing automatic CNN design algorithms in accuracy and efficiency.

Achieves comparable accuracy to manually-designed CNNs with less computational resources.

Validated on multiple benchmark datasets with superior results.

Abstract

Convolutional Neural Networks (CNNs) have gained a remarkable success on many image classification tasks in recent years. However, the performance of CNNs highly relies upon their architectures. For most state-of-the-art CNNs, their architectures are often manually-designed with expertise in both CNNs and the investigated problems. Therefore, it is difficult for users, who have no extended expertise in CNNs, to design optimal CNN architectures for their own image classification problems of interest. In this paper, we propose an automatic CNN architecture design method by using genetic algorithms, to effectively address the image classification tasks. The most merit of the proposed algorithm remains in its "automatic" characteristic that users do not need domain knowledge of CNNs when using the proposed algorithm, while they can still obtain a promising CNN architecture for the given images. The proposed algorithm is validated on widely used benchmark image classification datasets, by comparing to the state-of-the-art peer competitors covering eight manually-designed CNNs, seven automatic+manually tuning and five automatic CNN architecture design algorithms. The experimental results indicate the proposed algorithm outperforms the existing automatic CNN architecture design algorithms in terms of classification accuracy, parameter numbers and consumed computational resources. The proposed algorithm also shows the very comparable classification accuracy to the best one from manually-designed and automatic+manually tuning CNNs, while consumes much less of computational resource.