RC-Net: A Convolutional Neural Network for Retinal Vessel Segmentation

TL;DR

RC-Net is a simplified, fully convolutional neural network designed for retinal vessel segmentation that achieves competitive performance with significantly fewer parameters by optimizing filter counts and minimizing pooling.

Contribution

The paper introduces RC-Net, a less complex CNN architecture that maintains high accuracy in retinal vessel segmentation while drastically reducing the number of trainable parameters.

Findings

RC-Net outperforms other methods on benchmark datasets.

RC-Net uses fewer parameters, reducing model complexity.

The approach maintains high segmentation accuracy with simplified architecture.

Abstract

Over recent years, increasingly complex approaches based on sophisticated convolutional neural network architectures have been slowly pushing performance on well-established benchmark datasets. In this paper, we take a step back to examine the real need for such complexity. We present RC-Net, a fully convolutional network, where the number of filters per layer is optimized to reduce feature overlapping and complexity. We also used skip connections to keep spatial information loss to a minimum by keeping the number of pooling operations in the network to a minimum. Two publicly available retinal vessel segmentation datasets were used in our experiments. In our experiments, RC-Net is quite competitive, outperforming alternatives vessels segmentation methods with two or even three orders of magnitude less trainable parameters.