CASIA OpenIR  > 模式识别国家重点实验室  > 模式分析与学习
Online and offline handwritten Chinese character recognition: A comprehensive study and new benchmark
Zhang, Xu-Yao1; Bengio, Yoshua2; Liu, Cheng-Lin1,3; Xu-Yao Zhang
AbstractRecent deep learning based methods have achieved the state-of-the-art performance for handwritten Chinese character recognition (HCCR) by learning discriminative representations directly from raw data. Nevertheless, we believe that the long-and-well investigated domain-specific knowledge should still help to boost the performance of HCCR. By integrating the traditional normalization-cooperated direction-decomposed feature map (directMap) with the deep convolutional neural network (convNet), we are able to obtain new highest accuracies for both online and offline HCCR on the ICDAR-2013 competition database. With this new framework, we can eliminate the needs for data augmentation and model ensemble, which are widely used in other systems to achieve their best results. This makes our framework to be efficient and effective for both training and testing. Furthermore, although directMap+ convNet can achieve the best results and surpass human-level performance, we show that writer adaptation in this case is still effective. A new adaptation layer is proposed to reduce the mismatch between training and test data on a particular source layer. The adaptation process can be efficiently and effectively implemented in an unsupervised manner. By adding the adaptation layer into the pre-trained convNet, it can adapt to the new handwriting styles of particular writers, and the recognition accuracy can be further improved consistently and significantly. This paper gives an overview and comparison of recent deep learning based approaches for HCCR, and also sets new benchmarks for both online and offline HCCR. (C) 2016 Elsevier Ltd. All rights reserved.
KeywordHandwriting Recognition Chinese Characters Online Offline Directional Feature Map Convolutional Neural Network Adaptation
WOS HeadingsScience & Technology ; Technology
Indexed BySCI
Funding OrganizationNational Basic Research Program of China (973 Program)(2012CB316302) ; National Natural Science Foundation of China (NSFC)(61403380) ; Strategic Priority Research Program of the Chinese Academy of Sciences(XDA06040102 ; XDB02060009)
WOS Research AreaComputer Science ; Engineering
WOS SubjectComputer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS IDWOS:000385899400027
Citation statistics
Cited Times:47[WOS]   [WOS Record]     [Related Records in WOS]
Document Type期刊论文
Corresponding AuthorXu-Yao Zhang
Affiliation1.Chinese Acad Sci, Inst Automat, NLPR, Beijing, Peoples R China
2.Univ Montreal, MILA, Montreal, PQ H3C 3J7, Canada
3.CAS Ctr Excellence Brain Sci & Intelligence Techn, Beijing, Peoples R China
Recommended Citation
GB/T 7714
Zhang, Xu-Yao,Bengio, Yoshua,Liu, Cheng-Lin,et al. Online and offline handwritten Chinese character recognition: A comprehensive study and new benchmark[J]. PATTERN RECOGNITION,2017,61(61):348-360.
APA Zhang, Xu-Yao,Bengio, Yoshua,Liu, Cheng-Lin,&Xu-Yao Zhang.(2017).Online and offline handwritten Chinese character recognition: A comprehensive study and new benchmark.PATTERN RECOGNITION,61(61),348-360.
MLA Zhang, Xu-Yao,et al."Online and offline handwritten Chinese character recognition: A comprehensive study and new benchmark".PATTERN RECOGNITION 61.61(2017):348-360.
Files in This Item: Download All
File Name/Size DocType Version Access License
1-s2.0-S003132031630(1637KB)期刊论文作者接受稿开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Zhang, Xu-Yao]'s Articles
[Bengio, Yoshua]'s Articles
[Liu, Cheng-Lin]'s Articles
Baidu academic
Similar articles in Baidu academic
[Zhang, Xu-Yao]'s Articles
[Bengio, Yoshua]'s Articles
[Liu, Cheng-Lin]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Zhang, Xu-Yao]'s Articles
[Bengio, Yoshua]'s Articles
[Liu, Cheng-Lin]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 1-s2.0-S0031320316302187-main.pdf
Format: Adobe PDF
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.