Bones Can't Be Triangles: Accurate and Efficient Vertebrae Keypoint Estimation through Collaborative Error Revision

TL;DR

This paper introduces KeyBot, a novel method for vertebrae keypoint estimation that automatically identifies and corrects typical errors, significantly improving accuracy and reducing user effort in interactive annotation tasks.

Contribution

KeyBot is the first approach to automatically correct typical errors in vertebrae keypoint estimation, leveraging simulated error training for improved accuracy and efficiency.

Findings

Outperforms existing methods on three public datasets

Achieves state-of-the-art accuracy in interactive vertebrae keypoint estimation

Reduces user workload through effective error correction

Abstract

Recent advances in interactive keypoint estimation methods have enhanced accuracy while minimizing user intervention. However, these methods require user input for error correction, which can be costly in vertebrae keypoint estimation where inaccurate keypoints are densely clustered or overlap. We introduce a novel approach, KeyBot, specifically designed to identify and correct significant and typical errors in existing models, akin to user revision. By characterizing typical error types and using simulated errors for training, KeyBot effectively corrects these errors and significantly reduces user workload. Comprehensive quantitative and qualitative evaluations on three public datasets confirm that KeyBot significantly outperforms existing methods, achieving state-of-the-art performance in interactive vertebrae keypoint estimation. The source code and demo video are available at: https://ts-kim.github.io/KeyBot/