%0 Conference Proceedings
%T HTR-Flor: a deep learning system for offline handwritten text recognition
%D 2020
%8 Nov. 7-10, 2020
%A Souza Neto, Arthur Flor de,
%A Bezerra, Byron Leite Dantas,
%A Toselli, Alejandro Hector,
%A Lima, Estanislau Baptista,
%@affiliation Universidade de Pernambuco
%@affiliation Universidade de Pernambuco
%@affiliation Universitat Politecnica de Valencia
%@affiliation Universidade de Pernambuco
%E Musse, Soraia Raupp,
%E Cesar Junior, Roberto Marcondes,
%E Pelechano, Nuria,
%E Wang, Zhangyang (Atlas),
%B Conference on Graphics, Patterns and Images, 33 (SIBGRAPI)
%C Virtual
%S Proceedings
%I IEEE Computer Society
%J Los Alamitos
%K Handwritten Text Recognition, Gated Convolutional Neural Networks, Gated CNN, Deep Neural Networks.
%X In recent years, Handwritten Text Recognition (HTR) has captured a lot of attention among the researchers of the computer vision community. Current state-of-the-art approaches for offline HTR are based on Convolutional Recurrent Neural Networks (CRNNs) excel at scene text recognition. Unfortunately, deep models such as CRNNs, Recurrent Neural Networks (RNNs) are likely to suffer from vanishing/exploding gradient problems when processing long text images, which are commonly found in scanned documents. Besides, they usually have millions of parameters which require huge amount of data, and computational resource. Recently, a new class of neural network architecture, called Gated Convolutional Neural Networks (Gated-CNN), has demonstrated potentials to complement CRNN methods in modeling. Therefore, in this paper, we present a new architecture for HTR, based on Gated-CNN, with fewer parameters and fewer layers, which is able to outperform the current state-of-the-art architectures for HTR. The experiment validates that the proposed model has statistically significant recognition results, surpassing previous HTR systems by an average of 33% over five important handwritten benchmark datasets. Moreover, the proposed model is able to achieve satisfactory recognition rates even in case of few training data. Finally, its compact architecture requires less computational resources, which can be applied for real-world applications that have hardware limitations, such as robots and smartphones.
%@language en
%3 PID6607213.pdf