2021 Chinese Classic Text OCR Challenge

2021 Chinese Classic Text OCR Challenge

Introduction

Chinese OCR is an important element for digital scholarship and it is the first step for research on Chinese Classics text analysis.  The Chinese University of Hong Kong Library now invites related scholars and institutions to join the first “2021 Chinese Classic Text OCR Challenge”.  From this online Challenge the Library hopes to compare the excellence of the automatic recognition of Chinese characters in classics. Researchers or the industry can also study and exchange in enhancing the technology on Chinese classic texts automatic recognition, which in return enhances and promotes discovery of Chinese materials in academia for research in different disciplines.

Rules & Regulations

Schedule

2021.2.8 (Mon)Announcement of Challenge details and starts of registration
2021.3.1 (Mon)Deadline for registration
2021.3.4 (Thu)Online Briefing Session
2021.3.17 – 30Online Challenge (1 challenge/day for 10 days)
2021.4.2 (Fri)Final result announcement to participating teams
2021.4.7 (Wed)Winning List Announcement in Results Page
2021.4.21 (Wed)Online Prize Ceremony and Winner Presentation

Eligibility and Registration 

We welcome teams or individuals from all over the world to join the Challenge. Please register at https://bit.ly/3rp1ULH.  Deadline for registration is 1 March 2021 (Mon).

Challenge Rules

During the Challenge period, contestants will have to process the image files released every working day and organizer submits and processes the image file(TIFF)through online user interface provided by contestants.  Daily results will be released with assigned team code.

Assessment Criteria

Accuracy on recognized texts/text sequence and processing time of the output text file.

Result Announcement & Prizes 

Daily results will be sent to individual teams and the Top 5 scores will be published in the Results Page with team code.  Winner list can be found in Results Page and the Champion Team will give a presentation in the online Prize Ceremony held at 10:00 a.m. on 21 April 2021 (Wed)

Prizes: 

Champion: HKD8,000 of e-voucher & e-badge for display
1st Runner up: HKD5,000 of e-voucher 
2nd Runner up: HKD3,000 of e-voucher 

Click to Results Page

Intellectual Property

The organiser reserves the ownership of the original source file and contestants will only be authorized to use the files during the Challenge.  They are not allowed to use the original source files for other purposes without the authorisation of the organiser. The copyright of the text files submitted by the contestants belongs to the organiser. 

Participants must ensure that the technology or programs used in the competition file are original by the participants or legally authorized by the owner of the intellectual property rights.  If any third party raises allegations of infringement of intellectual property rights or the legal irregularities, the participants assume all legal responsibility.

Acknowledgement

Special thanks to Prof. Lai Chi-tim, Centre for Studies of Daoist Culture of CUHK in providing the Challenge text files.

Participating Teams

Below are the participating teams who have shared their contact for further exchange of the Chinese classic text OCR technology (list by random):

台灣中央研究院數位文化中心
Contact: 王祥安
Email: sawang @ gate.sinica.edu.tw
Website: https://www.ascdc.sinica.edu.tw/

陕西师范大学历史文化学院
Contact: 张光伟
Email: zhangguangwei @ snnu.edu.cn
Website: http://his.snnu.edu.cn/info/1069/8258.htm

深延科技(北京)有限公司
Contact: 陈彬、徐振宇
Email: chenb1 @ deepblueai.com / xuzy @ deepblueai.com
Website: http://shenyanai.com/

台灣國立台灣大學工程科學與海洋工程所 CKLab
Contact: 黃亁綱教授
Team members: 王婷、李家名、陳昱安
Email: ckhuang @ ntu.edu.tw

西安文数保科技有限公司
Contact:  闫升
Email: sheng.yan @ hengdawb.com

Enquiries

Read our FAQ page for any questions related to the Challenge.  

Please email to dslab@lib.cuhk.edu.hk for enquiries of the Challenge.