Full-scale Representation Guided Network for Retinal Vessel Segmentation

TL;DR

This paper introduces FSG-Net, a flexible retinal vessel segmentation model that uses full-scale structural features and attention-guided filtering to achieve state-of-the-art performance with a compact architecture.

Contribution

The paper proposes a novel Full-Scale Guided Network with an attention-guided filter and flexible architecture compatible with U-Net variants, improving retinal vessel segmentation.

Findings

FSG-Net achieves competitive performance with SOTA methods.

Each component of FSG-Net significantly improves segmentation accuracy.

The model is scalable and adaptable across different datasets.

Abstract

The U-Net architecture and its variants have remained state-of-the-art (SOTA) for retinal vessel segmentation over the past decade. In this study, we introduce a Full-Scale Guided Network (FSG-Net), where a novel feature representation module using modernized convolution blocks effectively captures full-scale structural information, while a guided convolution block subsequently refines this information. Specifically, we introduce an attention-guided filter within the guided convolution block, leveraging its similarity to unsharp masking to enhance fine vascular structures. Passing full-scale information to the attention block facilitates the generation of more contextually relevant attention maps, which are then passed to the attention-guided filter, providing further refinement to the segmentation performance. The structure preceding the guided convolution block can be replaced by any U-Net variant, ensuring flexibility and scalability across various segmentation tasks. For a fair comparison, we re-implemented recent studies available in public repositories to evaluate their scalability and reproducibility. Our experiments demonstrate that, despite its compact architecture, FSG-Net delivers performance competitive with SOTA methods across multiple public datasets. Ablation studies further demonstrate that each proposed component meaningfully contributes to this competitive performance. Our code is available on https://github.com/ZombaSY/FSG-Net-pytorch.