Hierarchical Document Parsing via Large Margin Feature Matching and Heuristics

TL;DR

This paper introduces a hierarchical document parsing method that combines large margin feature matching with heuristics, achieving high accuracy and efficiency in structure recognition tasks.

Contribution

It presents a novel integration of large margin loss and heuristic rules for improved hierarchical document parsing performance.

Findings

Achieved 0.98904 accuracy on the private leaderboard.

Outperformed existing methods in document structure parsing.

Demonstrated effectiveness of combining deep learning with heuristics.

Abstract

We present our solution to the AAAI-25 VRD-IU challenge, achieving first place in the competition. Our approach integrates large margin loss for improved feature discrimination and employs heuristic rules to refine hierarchical relationships. By combining a deep learning-based matching strategy with greedy algorithms, we achieve a significant boost in accuracy while maintaining computational efficiency. Our method attains an accuracy of 0.98904 on the private leaderboard, demonstrating its effectiveness in document structure parsing. Source codes are publicly available at https://github.com/ffyyytt/VRUID-AAAI-DAKiet