1st Place Solution for ICDAR 2021 Competition on Mathematical Formula Detection

Abstract

In this technical report, we present our 1st place solution for the ICDAR 2021 competition on mathematical formula detection (MFD). The MFD task has three key challenges including a large scale span, large variation of the ratio between height and width, and rich character set and mathematical expressions. Considering these challenges, we used Generalized Focal Loss (GFL), an anchor-free method, instead of the anchor-based method, and prove the Adaptive Training Sampling Strategy (ATSS) and proper Feature Pyramid Network (FPN) can well solve the important issue of scale variation. Meanwhile, we also found some tricks, e.g., Deformable Convolution Network (DCN), SyncBN, and Weighted Box Fusion (WBF), were effective in MFD task. Our proposed method ranked 1st in the final 15 teams.

arXiv comment1st Place Solution for ICDAR 2021 Competition on Mathematical Formula Detection. http://transcriptorium.eu/~htrcontest/MathsICDAR2021/

References in corpus (9)

• Adam: A Method for Stochastic Optimization
• FCOS: Fully Convolutional One-Stage Object Detection
• MMDetection: Open MMLab Detection Toolbox and Benchmark
• Weighted boxes fusion: Ensembling boxes from different object detection models
• Deformable Convolutional Networks
• Lookahead Optimizer: k steps forward, 1 step back
• Gradient Centralization: A New Optimization Technique for Deep Neural Networks
• PingAn-VCGroup's Solution for ICDAR 2021 Competition on Scientific Literature Parsing Task B: Table Recognition to HTML
• PingAn-VCGroup's Solution for ICDAR 2021 Competition on Scientific Table Image Recognition to Latex