Multi-Resolution Feature Stem for Diabetic Retinopathy lesion segmentation
2026-07-09 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors studied how different image sizes affect identifying diabetic eye problems in pictures. They found that using bigger images helps spot tiny issues but makes it harder to detect larger problems. To fix this, they created a new method that looks at multiple image sizes at the same time, capturing both small and big details. Their work shows that bigger pictures aren’t always better and offers a smart way to balance this trade-off.
Diabetic RetinopathyLesion SegmentationDeep LearningInput ResolutionU-NetUNet++Vision TransformersDeepLabV3+Multi-ResolutionImage Segmentation
Authors
Indranil Dutta, Taehee Jeong
Abstract
Diabetic Retinopathy (DR) is a leading cause of preventable blindness worldwide, requiring automated lesion segmentation using deep learning models for early detection and monitoring. However, DR lesions vary dramatically in size from tiny microaneurysms to large hemorrhages and exudates. This variability creates conflicting demands on the model architecture and input resolution, posing a challenge for effective design. This work investigates the impact of input resolution on different lesion types. Through systematic experimentation with multiple architectures (U-Net, UNet++, Vision Transformers, DeepLabV3+) at $512 \times 512$ and $1024 \times 1024$ resolutions, we identify a critical, counter-intuitive phenomenon where increasing input resolution has opposing effects on different lesion types. We demonstrate that while higher resolution is essential for resolving fine-grained microaneurysms, it can unexpectedly degrade performance on larger hemorrhages. This finding challenges the common assumption that higher resolution is uniformly beneficial. To address this, we propose a novel Multi-Resolution Feature Stem, an input-level pyramid integrated with a UNet++ backbone. This architecture processes multiple scales in parallel, capturing fine-grained details without sacrificing contextual information. This work contributes crucial empirical evidence of this complex, resolution-dependent behavior and a practical, parameter-efficient architecture that successfully resolves this trade-off.