Hierarchical Document Parsing via Large Margin Feature Matching and Heuristics
This work addresses document structure parsing for AI applications, but it is incremental as it combines existing deep learning and heuristic methods.
The paper tackled the problem of hierarchical document parsing by integrating large margin loss for feature discrimination and heuristic rules for refining relationships, achieving an accuracy of 0.98904 on the private leaderboard in the AAAI-25 VRD-IU challenge.
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