Scale-Aware Self-Supervised Learning for Segmentation of Small and Sparse Structures

TL;DR

This paper introduces a scale-aware self-supervised learning method that improves segmentation of small, sparse, and irregular structures by focusing on fine-scale features during pretraining, demonstrating significant gains in seismic and neuroimaging tasks.

Contribution

It proposes a novel scale-aware SSL approach with small-window cropping, tailored for small and sparse structures, and validates its effectiveness across different scientific imaging domains.

Findings

Up to 13% accuracy improvement in fault segmentation

Up to 5% accuracy improvement in cell delineation

Scale-aware SSL benefits are most pronounced for small, sparse objects

Abstract

Self-supervised learning (SSL) has emerged as a powerful strategy for representation learning under limited annotation regimes, yet its effectiveness remains highly sensitive to many factors, especially the nature of the target task. In segmentation, existing pipelines are typically tuned to large, homogeneous regions, but their performance drops when objects are small, sparse, or locally irregular. In this work, we propose a scale-aware SSL adaptation that integrates small-window cropping into the augmentation pipeline, zooming in on fine-scale structures during pretraining. We evaluate this approach across two domains with markedly different data modalities: seismic imaging, where the goal is to segment sparse faults, and neuroimaging, where the task is to delineate small cellular structures. In both settings, our method yields consistent improvements over standard and state-of-the-art baselines under label constraints, improving accuracy by up to 13% for fault segmentation and 5% for cell delineation. In contrast, large-scale features such as seismic facies or tissue regions see little benefit, underscoring that the value of SSL depends critically on the scale of the target objects. Our findings highlight the need to align SSL design with object size and sparsity, offering a general principle for buil ding more effective representation learning pipelines across scientific imaging domains.