Skeleton-Guided Progressive Test-Time Adaptation for Thin Curvilinear Structures

2026-10-07

AI Summary

This study addresses the segmentation failure of fine curvilinear structures caused by topological fragility under cross-modal domain shifts. To this end, we propose a skeleton-guided progressive test-time adaptation method. The core innovations include a novel progressive batch normalization scheduling strategy and a skeleton-consensus recall mechanism that updates only affine parameters, explicitly constraining and preserving structural connectivity through dynamic statistic transfer. Experimental results demonstrate that the proposed approach significantly improves topological connectivity metrics in extreme cross-modal scenarios, substantially outperforming existing test-time adaptation methods.

Abstract

Accurate segmentation of thin curvilinear structures is vital for various real-world applications, from vessel analysis to road extraction. Yet their intricate geometry makes even minor pixel-wise errors enough to break the global topology, and this structural fragility turns severe domain shifts into catastrophic failures. The difficulty is most acute under cross-modality gaps, where the imaging process itself differs fundamentally between source and target. While test-time adaptation (TTA) offers a practical source-free remedy, existing methods adapt feature statistics and confidence, neither of which constrains connectivity, and thus degrade under such extreme gaps. To address this, we propose Skeleton-Guided Progressive Test-Time Adaptation (SGP-TTA). Progressive Batch Normalization (ProgBN) shifts normalization from frozen source statistics toward current target estimates under a sample-count schedule, so that the source-target balance follows the stage of adaptation rather than a fixed coefficient. Consensus Skeleton Recall (CSR) then derives a structural target from geometrically aligned multi-view predictions and updates only the BN affine parameters to preserve connected structures. Extensive experiments show that SGP-TTA consistently outperforms existing TTA methods in topological connectivity, with the largest margins under cross-modality shift. The project page is available at https://boa-jang.github.io/SGP-TTA.

Problem

• thin curvilinear structures
• test-time adaptation
• domain shift
• cross-modality
• topological connectivity

Innovation

• Test-Time Adaptation
• Skeleton-Guided
• Progressive Batch Normalization
• Consensus Skeleton Recall
• Thin Curvilinear Structures