SA-UNetv2: Rethinking Spatial Attention U-Net for Retinal Vessel Segmentation
Abstract
Retinal vessel segmentation is essential for early diagnosis of diseases such as diabetic retinopathy, hypertension, and neurodegenerative disorders. Although SA-UNet introduces spatial attention in the bottleneck, it underuses attention in skip connections and does not address the severe foreground-background imbalance. We propose SA-UNetv2, a lightweight model that injects cross-scale spatial attention into all skip connections to strengthen multi-scale feature fusion and adopts a weighted Binary Cross-Entropy (BCE) plus Matthews Correlation Coefficient (MCC) loss to improve robustness to class imbalance. On the public DRIVE and STARE datasets, SA-UNetv2 achieves state-of-the-art performance with only 1.2MB memory and 0.26M parameters (less than 50% of SA-UNet), and 1 second CPU inference on 592 x 592 x 3 images, demonstrating strong efficiency and deployability in resource-constrained, CPU-only settings.
Cite
@article{arxiv.2509.11774,
title = {SA-UNetv2: Rethinking Spatial Attention U-Net for Retinal Vessel Segmentation},
author = {Changlu Guo and Anders Nymark Christensen and Anders Bjorholm Dahl and Yugen Yi and Morten Rieger Hannemose},
journal= {arXiv preprint arXiv:2509.11774},
year = {2025}
}
Comments
The code is available at github.com/clguo/SA-UNetv2