Fusion of ConvNeXt-Tiny and Swin-Tiny backbones: a comparative analysis for diabetic retinopathy classification.

In this study, we perform a controlled comparative evaluation of ConvNeXt-Tiny, Swin-Tiny and their feature fusion for diabetic retinopathy classification using the Asia Pacific Tele-Ophthalmology Society (Asia Pacific Tele-Ophthalmology Society) 2019 dataset.

All models were initialized with weights pre-trained on ImageNet-1K and evaluated with two transfer learning strategies: direct fine-tuning on Asia Pacific Tele-Ophthalmology Society 2019, and EyePACS-based domain adaptation with task-specific fine-tuning.