Mathematical Variable Detection based on Convolutional Neural Network and Support Vector Machine
Bui Hai Phong, Thang Manh Hoang, Thi‐Lan Le · 2019
Mathematical expression detection in scientific documents is a prerequisite step for developing a mathematical retrieval system that has attracted many researches recently. In the detecting process, a challenging issue is the detection of variable. The similar properties of variable and narrative text cause many errors in the detection in existing approaches. In the paper, pre-trained deep Convolutional Neural Networks (CNN) are employed and optimized to automatically extract visual features of images and Support Vector Machine (SVM) is used to improve the accuracy of the detection. The accuracy of 99.5% is achieved for the detection of variable in inline expressions in document images in public benchmark datasets. The performance comparison with traditional methods demonstrates the effectiveness of the proposed method.