Improving Faster R-CNN Framework for Multiscale Chinese Character Detection and Localization
Summary :
Faster R-CNN uses a region proposal network which consists of a single scale convolution filter and fully connected networks to localize detected regions. However, using a single scale filter is not enough to detect full regions of characters. In this letter, we propose a simple but effective way, i.e., utilizing variously sized convolution filters, to accurately detect Chinese characters of multiple scales in documents. We experimentally verified that our method improved IoU by 4% and detection rate by 3% than the previous single scale Faster R-CNN method.
- Publication
- IEICE TRANSACTIONS on Information Vol.E103-D No.7 pp.1777-1781
- Publication Date
- 2020/07/01
- Publicized
- 2020/04/06
- Online ISSN
- 1745-1361
- DOI
- 10.1587/transinf.2019EDL8217
- Type of Manuscript
- LETTER
- Category
- Pattern Recognition
Authors
Minseong KIM
Yeungnam University
Hyun-Chul CHOI
Yeungnam University
Keyword
Latest Issue
Copyrights notice of machine-translated contents
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Cite this
Copy
Minseong KIM, Hyun-Chul CHOI, "Improving Faster R-CNN Framework for Multiscale Chinese Character Detection and Localization" in IEICE TRANSACTIONS on Information,
vol. E103-D, no. 7, pp. 1777-1781, July 2020, doi: 10.1587/transinf.2019EDL8217.
Abstract: Faster R-CNN uses a region proposal network which consists of a single scale convolution filter and fully connected networks to localize detected regions. However, using a single scale filter is not enough to detect full regions of characters. In this letter, we propose a simple but effective way, i.e., utilizing variously sized convolution filters, to accurately detect Chinese characters of multiple scales in documents. We experimentally verified that our method improved IoU by 4% and detection rate by 3% than the previous single scale Faster R-CNN method.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2019EDL8217/_p
Copy
@ARTICLE{e103-d_7_1777,
author={Minseong KIM, Hyun-Chul CHOI, },
journal={IEICE TRANSACTIONS on Information},
title={Improving Faster R-CNN Framework for Multiscale Chinese Character Detection and Localization},
year={2020},
volume={E103-D},
number={7},
pages={1777-1781},
abstract={Faster R-CNN uses a region proposal network which consists of a single scale convolution filter and fully connected networks to localize detected regions. However, using a single scale filter is not enough to detect full regions of characters. In this letter, we propose a simple but effective way, i.e., utilizing variously sized convolution filters, to accurately detect Chinese characters of multiple scales in documents. We experimentally verified that our method improved IoU by 4% and detection rate by 3% than the previous single scale Faster R-CNN method.},
keywords={},
doi={10.1587/transinf.2019EDL8217},
ISSN={1745-1361},
month={July},}
Copy
TY - JOUR
TI - Improving Faster R-CNN Framework for Multiscale Chinese Character Detection and Localization
T2 - IEICE TRANSACTIONS on Information
SP - 1777
EP - 1781
AU - Minseong KIM
AU - Hyun-Chul CHOI
PY - 2020
DO - 10.1587/transinf.2019EDL8217
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 7
JA - IEICE TRANSACTIONS on Information
Y1 - July 2020
AB - Faster R-CNN uses a region proposal network which consists of a single scale convolution filter and fully connected networks to localize detected regions. However, using a single scale filter is not enough to detect full regions of characters. In this letter, we propose a simple but effective way, i.e., utilizing variously sized convolution filters, to accurately detect Chinese characters of multiple scales in documents. We experimentally verified that our method improved IoU by 4% and detection rate by 3% than the previous single scale Faster R-CNN method.
ER -
English
