Abstract
Automatic handwriting recognition systems are of interest for academic research fields and for commercial applications. Recent advances in deep learning techniques have shown dramatic improvement in relation to classic computer vision problems, especially in Handwritten Text Recognition (HTR). However, several approaches try to solve the problem of deep learning applied to Handwritten Digit String Recognition (HDSR), where it has to deal with the low number of trainable data, while learning to ignore any writing symbol around the digits (noise). In this context, we present a new optical model architecture (Gated-CNN-BGRU), based on HTR workflow, applied to HDSR. The International Conference on Frontiers of Handwriting Recognition (ICFHR) 2014 competition on HDSR were used as baselines to evaluate the effectiveness of our proposal, whose metrics, datasets and recognition methods were adopted for fair comparison. Furthermore, we also use a private dataset (Brazilian Bank Check - Courtesy Amount Recognition), and 11 different approaches from the state-of-the-art in HDSR, as well as 2 optical models from the state-of-the-art in HTR. Finally, the proposed optical model demonstrated robustness even with low data volume (126 trainable data, for example), surpassing the results of existing methods with an average precision of 96.50%, which is equivalent to an average percentage of improvement of 3.74 points compared to the state-of-the-art in HDSR. In addition, the result stands out in the competition’s CVL HDS set, where the proposed optical model achieved a precision of 93.54%, while the best result so far had been from Beijing group (from the competition itself), with 85.29%.
Highlights
H ANDWRITTEN Digit String Recognition (HDSR) has attracted intense attention in recent years as an academic research topic, due to its vast applications in industry [1]
Automatic Handwritten Digit String Recognition (HDSR) systems are very helpful in offline handwriting recognition applications, where the recognition is performed through images
We propose a new Gated Convolutional Recurrent Neural Network using Connectionist Temporal Classification (CTC) (Gated-Convolutional Recurrent Neural Networks (CRNN)-CTC) as optical model for HDSR
Summary
H ANDWRITTEN Digit String Recognition (HDSR) has attracted intense attention in recent years as an academic research topic, due to its vast applications in industry [1]. The HDSR task has been a subject of study for many years, and a variety of methods and algorithms have been proposed to improve recognition precision [3]. Traditional approaches are based on over-segmentation strategy, to classify the detected digits. This kind of methods faced many problems in practice, such as various styles of handwriting, connected characters and background noise [4]. End-to-end systems are used to deal with these problems, where neural network models extract features from images and recognize the digits, without prior detection [5]–[8]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
