Link prediction Graph Neural Networks for structure recognition of Handwritten Mathematical Expressions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Nguyen, Cuong Tuan, Nguyen, Ngoc Tuan, Dao, Triet Hoang Minh, Nhat, Huy Minh, Dinh, Huy Truong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912686836547584
author Nguyen, Cuong Tuan
Nguyen, Ngoc Tuan
Dao, Triet Hoang Minh
Nhat, Huy Minh
Dinh, Huy Truong
author_facet Nguyen, Cuong Tuan
Nguyen, Ngoc Tuan
Dao, Triet Hoang Minh
Nhat, Huy Minh
Dinh, Huy Truong
contents We propose a Graph Neural Network (GNN)-based approach for Handwritten Mathematical Expression (HME) recognition by modeling HMEs as graphs, where nodes represent symbols and edges capture spatial dependencies. A deep BLSTM network is used for symbol segmentation, recognition, and spatial relation classification, forming an initial primitive graph. A 2D-CFG parser then generates all possible spatial relations, while the GNN-based link prediction model refines the structure by removing unnecessary connections, ultimately forming the Symbol Label Graph. Experimental results demonstrate the effectiveness of our approach, showing promising performance in HME structure recognition.
format Preprint
id arxiv_https___arxiv_org_abs_2511_02288
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Link prediction Graph Neural Networks for structure recognition of Handwritten Mathematical Expressions
Nguyen, Cuong Tuan
Nguyen, Ngoc Tuan
Dao, Triet Hoang Minh
Nhat, Huy Minh
Dinh, Huy Truong
Computer Vision and Pattern Recognition
Computation and Language
We propose a Graph Neural Network (GNN)-based approach for Handwritten Mathematical Expression (HME) recognition by modeling HMEs as graphs, where nodes represent symbols and edges capture spatial dependencies. A deep BLSTM network is used for symbol segmentation, recognition, and spatial relation classification, forming an initial primitive graph. A 2D-CFG parser then generates all possible spatial relations, while the GNN-based link prediction model refines the structure by removing unnecessary connections, ultimately forming the Symbol Label Graph. Experimental results demonstrate the effectiveness of our approach, showing promising performance in HME structure recognition.
title Link prediction Graph Neural Networks for structure recognition of Handwritten Mathematical Expressions
topic Computer Vision and Pattern Recognition
Computation and Language
url https://arxiv.org/abs/2511.02288