Improving Offline Handwritten Chinese Character Recognition by Iterative Refinement

Xiao Yang, Dafang He, Zihan Zhou, Daniel Kifer, Clyde Lee Giles

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

We present an iterative refinement module that can be applied to the output feature maps of any existing convolutional neural networks in order to further improve classification accuracy. The proposed module, implemented by an attention-based recurrent neural network, can iteratively use its previous predictions to update attention and thereafter refine current predictions. In this way, the model is able to focus on a sub-region of input images to distinguish visually similar characters (see Figure 1 for an example). We evaluate its effectiveness on handwritten Chinese character recognition (HCCR) task and observe significant performance gain. HCCR task is challenging due to large number of classes and small differences between certain characters. To overcome these difficulties, we further propose a novel convolutional architecture that utilizes both low-level visual cues and high-level structural information. Together with the proposed iterative refinement module, our approach achieves an accuracy of 97.37%, outperforming previous methods that use raw images as input on ICDAR-2013 dataset [1].

Original languageEnglish (US)
Title of host publicationProceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017
PublisherIEEE Computer Society
Pages5-10
Number of pages6
ISBN (Electronic)9781538635865
DOIs
StatePublished - Jan 25 2018
Event14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017 - Kyoto, Japan
Duration: Nov 9 2017Nov 15 2017

Publication series

NameProceedings of the International Conference on Document Analysis and Recognition, ICDAR
Volume1
ISSN (Print)1520-5363

Other

Other14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017
CountryJapan
CityKyoto
Period11/9/1711/15/17

Fingerprint

Character recognition
Recurrent neural networks
Neural networks

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Cite this

Yang, X., He, D., Zhou, Z., Kifer, D., & Giles, C. L. (2018). Improving Offline Handwritten Chinese Character Recognition by Iterative Refinement. In Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017 (pp. 5-10). (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR; Vol. 1). IEEE Computer Society. https://doi.org/10.1109/ICDAR.2017.11
Yang, Xiao ; He, Dafang ; Zhou, Zihan ; Kifer, Daniel ; Giles, Clyde Lee. / Improving Offline Handwritten Chinese Character Recognition by Iterative Refinement. Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017. IEEE Computer Society, 2018. pp. 5-10 (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR).
@inproceedings{14b6a3bcea66469c96130fb7a4947670,
title = "Improving Offline Handwritten Chinese Character Recognition by Iterative Refinement",
abstract = "We present an iterative refinement module that can be applied to the output feature maps of any existing convolutional neural networks in order to further improve classification accuracy. The proposed module, implemented by an attention-based recurrent neural network, can iteratively use its previous predictions to update attention and thereafter refine current predictions. In this way, the model is able to focus on a sub-region of input images to distinguish visually similar characters (see Figure 1 for an example). We evaluate its effectiveness on handwritten Chinese character recognition (HCCR) task and observe significant performance gain. HCCR task is challenging due to large number of classes and small differences between certain characters. To overcome these difficulties, we further propose a novel convolutional architecture that utilizes both low-level visual cues and high-level structural information. Together with the proposed iterative refinement module, our approach achieves an accuracy of 97.37{\%}, outperforming previous methods that use raw images as input on ICDAR-2013 dataset [1].",
author = "Xiao Yang and Dafang He and Zihan Zhou and Daniel Kifer and Giles, {Clyde Lee}",
year = "2018",
month = "1",
day = "25",
doi = "10.1109/ICDAR.2017.11",
language = "English (US)",
series = "Proceedings of the International Conference on Document Analysis and Recognition, ICDAR",
publisher = "IEEE Computer Society",
pages = "5--10",
booktitle = "Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017",
address = "United States",

}

Yang, X, He, D, Zhou, Z, Kifer, D & Giles, CL 2018, Improving Offline Handwritten Chinese Character Recognition by Iterative Refinement. in Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017. Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, vol. 1, IEEE Computer Society, pp. 5-10, 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017, Kyoto, Japan, 11/9/17. https://doi.org/10.1109/ICDAR.2017.11

Improving Offline Handwritten Chinese Character Recognition by Iterative Refinement. / Yang, Xiao; He, Dafang; Zhou, Zihan; Kifer, Daniel; Giles, Clyde Lee.

Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017. IEEE Computer Society, 2018. p. 5-10 (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR; Vol. 1).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Improving Offline Handwritten Chinese Character Recognition by Iterative Refinement

AU - Yang, Xiao

AU - He, Dafang

AU - Zhou, Zihan

AU - Kifer, Daniel

AU - Giles, Clyde Lee

PY - 2018/1/25

Y1 - 2018/1/25

N2 - We present an iterative refinement module that can be applied to the output feature maps of any existing convolutional neural networks in order to further improve classification accuracy. The proposed module, implemented by an attention-based recurrent neural network, can iteratively use its previous predictions to update attention and thereafter refine current predictions. In this way, the model is able to focus on a sub-region of input images to distinguish visually similar characters (see Figure 1 for an example). We evaluate its effectiveness on handwritten Chinese character recognition (HCCR) task and observe significant performance gain. HCCR task is challenging due to large number of classes and small differences between certain characters. To overcome these difficulties, we further propose a novel convolutional architecture that utilizes both low-level visual cues and high-level structural information. Together with the proposed iterative refinement module, our approach achieves an accuracy of 97.37%, outperforming previous methods that use raw images as input on ICDAR-2013 dataset [1].

AB - We present an iterative refinement module that can be applied to the output feature maps of any existing convolutional neural networks in order to further improve classification accuracy. The proposed module, implemented by an attention-based recurrent neural network, can iteratively use its previous predictions to update attention and thereafter refine current predictions. In this way, the model is able to focus on a sub-region of input images to distinguish visually similar characters (see Figure 1 for an example). We evaluate its effectiveness on handwritten Chinese character recognition (HCCR) task and observe significant performance gain. HCCR task is challenging due to large number of classes and small differences between certain characters. To overcome these difficulties, we further propose a novel convolutional architecture that utilizes both low-level visual cues and high-level structural information. Together with the proposed iterative refinement module, our approach achieves an accuracy of 97.37%, outperforming previous methods that use raw images as input on ICDAR-2013 dataset [1].

UR - http://www.scopus.com/inward/record.url?scp=85045202503&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85045202503&partnerID=8YFLogxK

U2 - 10.1109/ICDAR.2017.11

DO - 10.1109/ICDAR.2017.11

M3 - Conference contribution

AN - SCOPUS:85045202503

T3 - Proceedings of the International Conference on Document Analysis and Recognition, ICDAR

SP - 5

EP - 10

BT - Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017

PB - IEEE Computer Society

ER -

Yang X, He D, Zhou Z, Kifer D, Giles CL. Improving Offline Handwritten Chinese Character Recognition by Iterative Refinement. In Proceedings - 14th IAPR International Conference on Document Analysis and Recognition, ICDAR 2017. IEEE Computer Society. 2018. p. 5-10. (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR). https://doi.org/10.1109/ICDAR.2017.11