Next Article in Journal
Concrete Delamination Depth Estimation Using a Noncontact MEMS Ultrasonic Sensor Array and an Optimization Approach
Next Article in Special Issue
Validated Filter-Based Photoreceptor Count Algorithm on Retinal Heidelberg High Magnification Module™ Images in Healthy and Pathological Conditions
Previous Article in Journal
Wind Turbine Prognosis Models Based on SCADA Data and Extreme Learning Machines
Previous Article in Special Issue
A New Gaze Estimation Method Based on Homography Transformation Derived from Geometric Relationship
Article

Leveraging the Generalization Ability of Deep Convolutional Neural Networks for Improving Classifiers for Color Fundus Photographs

1
VUNO Inc., Seoul 06709, Korea
2
Genesis Lab Inc., Seoul 04538, Korea
*
Author to whom correspondence should be addressed.
Appl. Sci. 2021, 11(2), 591; https://doi.org/10.3390/app11020591
Received: 30 November 2020 / Revised: 31 December 2020 / Accepted: 6 January 2021 / Published: 9 January 2021
(This article belongs to the Special Issue Biomedical Engineering Applications in Vision Science)
Deep learning demands a large amount of annotated data, and the annotation task is often crowdsourced for economic efficiency. When the annotation task is delegated to non-experts, the dataset may contain data with inaccurate labels. Noisy labels not only yield classification models with sub-optimal performance, but may also impede their optimization dynamics. In this work, we propose exploiting the pattern recognition capacity of deep convolutional neural networks to filter out supposedly mislabeled cases while training. We suggest a training method that references softmax outputs to judge the correctness of the given labels. This approach achieved outstanding performance compared to the existing methods in various noise settings on a large-scale dataset (Kaggle 2015 Diabetic Retinopathy). Furthermore, we demonstrate a method mining positive cases from a pool of unlabeled images by exploiting the generalization ability. With this method, we won first place on the offsite validation dataset in pathological myopia classification challenge (PALM), achieving the AUROC of 0.9993 in the final submission. Source codes are publicly available. View Full-Text
Keywords: diabetic retinopathy; pathologic myopia; classification; retinal image analysis; fundus image; deep convolutional neural network; Semi-supervised learning diabetic retinopathy; pathologic myopia; classification; retinal image analysis; fundus image; deep convolutional neural network; Semi-supervised learning
Show Figures

Figure 1

MDPI and ACS Style

Son, J.; Kim, J.; Kong, S.T.; Jung, K.-H. Leveraging the Generalization Ability of Deep Convolutional Neural Networks for Improving Classifiers for Color Fundus Photographs. Appl. Sci. 2021, 11, 591. https://doi.org/10.3390/app11020591

AMA Style

Son J, Kim J, Kong ST, Jung K-H. Leveraging the Generalization Ability of Deep Convolutional Neural Networks for Improving Classifiers for Color Fundus Photographs. Applied Sciences. 2021; 11(2):591. https://doi.org/10.3390/app11020591

Chicago/Turabian Style

Son, Jaemin, Jaeyoung Kim, Seo T. Kong, and Kyu-Hwan Jung. 2021. "Leveraging the Generalization Ability of Deep Convolutional Neural Networks for Improving Classifiers for Color Fundus Photographs" Applied Sciences 11, no. 2: 591. https://doi.org/10.3390/app11020591

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop