DCML: Deep contrastive mutual learning for COVID-19 recognition. (August 2022)

• Record Type: Journal Article
• Authors: Zhang, Hongbin; Liang, Weinan; Li, Chuanxiu; Xiong, Qipeng; Shi, Haowei; Hu, Lang; Li, Guangli
• Is Part Of: Biomedical signal processing and control. Volume 77(2022)
• Publication Date: 2022-08
• Subjects: COVID-19 recognition -- Deep mutual learning -- Contrastive learning -- Fast AutoAugment -- Adaptive model fusion; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.103770
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: We propose a novel model, called DCML, for effective COVID-19 recognition. We propose a new adaptive feature fusion method. DCML tries to imitate practical diagnosis scenarios as much as possible. DCML is a good prototype combining contrastive learning and mutual learning. DCML is effective, efficient, lightweight, and robust. Abstract: COVID-19 is a form of disease triggered by a new strain of coronavirus. Automatic COVID-19 recognition using computer-aided methods is beneficial for speeding up diagnosis efficiency. Current researches usually focus on a deeper or wider neural network for COVID-19 recognition. And the implicit contrastive relationship between different samples has not been fully explored. To address these problems, we propose a novel model, called deep contrastive mutual learning (DCML), to diagnose COVID-19 more effectively. A multi-way data augmentation strategy based on Fast AutoAugment (FAA) was employed to enrich the original training dataset, which helps reduce the risk of overfitting. Then, we incorporated the popular contrastive learning idea into the conventional deep mutual learning (DML) framework to mine the relationship between diverse samples and created more discriminative image features through a new adaptive model fusion method. Experimental results on three public datasets demonstrate that the DCML model outperforms other state-of-the-art baselines. More importantly, DCML is easier to reproduce and relatively efficient, strengthening its high practicality.