Proposal-Free Volumetric Instance Segmentation from Latent Single-Instance Masks

TL;DR

This paper presents a proposal-free, memory-efficient volumetric instance segmentation method that predicts all masks simultaneously using a signed graph partitioning approach, demonstrating competitive results on neuron segmentation benchmarks.

Contribution

The method introduces a novel proposal-free approach that predicts all instance masks concurrently from a shared latent space, improving robustness and efficiency for large volumetric images.

Findings

Achieves competitive scores on CREMI 2016 neuron segmentation benchmark.

Concurrently predicts all masks, resolving conflicts across the entire image.

Uses a low-dimensional latent representation for memory efficiency.

Abstract

This work introduces a new proposal-free instance segmentation method that builds on single-instance segmentation masks predicted across the entire image in a sliding window style. In contrast to related approaches, our method concurrently predicts all masks, one for each pixel, and thus resolves any conflict jointly across the entire image. Specifically, predictions from overlapping masks are combined into edge weights of a signed graph that is subsequently partitioned to obtain all final instances concurrently. The result is a parameter-free method that is strongly robust to noise and prioritizes predictions with the highest consensus across overlapping masks. All masks are decoded from a low dimensional latent representation, which results in great memory savings strictly required for applications to large volumetric images. We test our method on the challenging CREMI 2016 neuron segmentation benchmark where it achieves competitive scores.