Adjusting the Ground Truth Annotations for Connectivity-Based Learning to Delineate

TL;DR

This paper presents a method that improves 3D structure delineation by jointly training deep networks and correcting annotation errors using deformable active contour models, enhancing accuracy despite annotation inaccuracies.

Contribution

Introduces a novel approach that explicitly models annotation inaccuracies as deformable contours, enabling joint training and correction of annotations during network learning.

Findings

Boosts deep network performance with inaccurate annotations

Joint training and correction improve delineation accuracy

Code released for reproducibility

Abstract

Deep learning-based approaches to delineating 3D structure depend on accurate annotations to train the networks. Yet, in practice, people, no matter how conscientious, have trouble precisely delineating in 3D and on a large scale, in part because the data is often hard to interpret visually and in part because the 3D interfaces are awkward to use. In this paper, we introduce a method that explicitly accounts for annotation inaccuracies. To this end, we treat the annotations as active contour models that can deform themselves while preserving their topology. This enables us to jointly train the network and correct potential errors in the original annotations. The result is an approach that boosts performance of deep networks trained with potentially inaccurate annotations. Code has been released at https://github.com/doruk-oner/AdjustingAnnotationswithSnakes.