ENSEMBLE APPROACH USING DEEP LEARNING AND TRANSFER LEARNING IN CLASSIFICATION OF DIABETIC RETINOPATHY
International Journal of Scientific Research in Engineering and Management · 2026-05 · 10 (5)
DOI: 10.55041/ijsrem63175 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Scientific Research in Engineering and Management, 2026-05. Verfasst von SHILPASHREE H C SHILPASHREE H C.
Abstract (Originalsprache)
Diabetic Retinopathy (DR) is a serious eye disease caused by long-term diabetes and may lead to blindness if not detected early.Deep Learning (DL) and Convolutional Neural Network (CNN) models are widely used for automated DR detection from retinal fundus images.In this study, an enhanced activation function is proposed to improve CNN performance by reducing training loss and computational complexity while increasing accuracy.Multiple public datasets, including DIARETDB0, DRIVE, CHASE, and Kaggle DR datasets, were used for evaluation.Among the tested models, ResNet-152 achieved the highest accuracy of 99.41% on the Kaggle dataset.The results demonstrate the effectiveness of the proposed approach for accurate and reliable DR diagnosis.
Zitieren
APA
C, S. H. C. S. H. (2026). ENSEMBLE APPROACH USING DEEP LEARNING AND TRANSFER LEARNING IN CLASSIFICATION OF DIABETIC RETINOPATHY. International Journal of Scientific Research in Engineering and Management, 10(5). https://doi.org/10.55041/ijsrem63175
Vancouver
C SHCSH. ENSEMBLE APPROACH USING DEEP LEARNING AND TRANSFER LEARNING IN CLASSIFICATION OF DIABETIC RETINOPATHY. International Journal of Scientific Research in Engineering and Management. 2026;10(5). doi:10.55041/ijsrem63175
BibTeX
@article{C2026,
author = {C, SHILPASHREE H C SHILPASHREE H},
title = {ENSEMBLE APPROACH USING DEEP LEARNING AND TRANSFER LEARNING IN CLASSIFICATION OF DIABETIC RETINOPATHY},
journal = {International Journal of Scientific Research in Engineering and Management},
year = {2026},
volume = {10},
number = {5},
doi = {10.55041/ijsrem63175}
}
EUVIMED