MultiStar: Instance Segmentation of Overlapping Objects with Star-Convex Polygons

TL;DR

MultiStar extends StarDist for instance segmentation of overlapping objects in biomedical images, effectively identifying overlaps to improve proposal sampling and segmentation accuracy with a simple, trainable network.

Contribution

It introduces a novel approach to handle overlapping objects by identifying overlap pixels, enhancing proposal sampling without significant overhead.

Findings

Effective segmentation of overlapping objects demonstrated on two datasets.

Maintains simplicity and ease of training in the network architecture.

Improved proposal sampling leads to better segmentation results.

Abstract

Instance segmentation of overlapping objects in biomedical images remains a largely unsolved problem. We take up this challenge and present MultiStar, an extension to the popular instance segmentation method StarDist. The key novelty of our method is that we identify pixels at which objects overlap and use this information to improve proposal sampling and to avoid suppressing proposals of truly overlapping objects. This allows us to apply the ideas of StarDist to images with overlapping objects, while incurring only a small overhead compared to the established method. MultiStar shows promising results on two datasets and has the advantage of using a simple and easy to train network architecture.