A multi-channel Inception V3 model with custom downscaling preprocessing outperforms prior methods on accuracy, sensitivity, and specificity when trained on the combined Kaggle and IDRiD diabetic retinopathy datasets.
Diagnostic methods i: Sensitivity, specificity, and other measures of accuracy
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Diabetic Retinopathy Classification using Downscaling Algorithms and Deep Learning
A multi-channel Inception V3 model with custom downscaling preprocessing outperforms prior methods on accuracy, sensitivity, and specificity when trained on the combined Kaggle and IDRiD diabetic retinopathy datasets.