Offline handwrittenformula recognition is a challenging task due to the variety of handwritten symbols and two-dimensional formula ***,the deep neural network recognizers based on the encoder-decoder frame-work have ...
详细信息
Offline handwrittenformula recognition is a challenging task due to the variety of handwritten symbols and two-dimensional formula ***,the deep neural network recognizers based on the encoder-decoder frame-work have achieved great improvements on this ***,the unsatisfactory recognition performance for formulas with long LTeX strings is one shortcoming of the existing ***,lacking sufficient training data also limits the capability of these *** this paper,we design a multimodal dependence attention(MDA)module to help the model learn visual and semantic dependencies among symbols in the same formula to improve the recognition perfor-mance of the formulas with long LTeX *** alleviate overfitting and further improve the recognition performance,we also propose a new dataset,handwritten formula image dataset(HFID),which contains 25620 handwrittenformulaimages collected from real *** conduct extensive experiments to demonstrate the effectiveness of our proposed MDA module and HFID dataset and achieve state-of-the-art performances,63.79%and 65.24%expression accuracy on CROHME 2014 and CROHME 2016,respectively.
暂无评论