top of page

Advance of Traditional Chinese Scene Text Recognition Competition

1.Purpose of the competition

The signs and advertisements around our lives now convey rich text messages to people. If we can recognize the text in the screen automatically, which is useful for scene understanding, smart city/transportation development, self-driving cars, assisting the visually impaired and foreign travelers. The Advance of Traditional Chinese Scene Text Recognition Competition aims to recognize Chinese characters in street scenes in Taiwan. We hope that this contest can enhance the technological development in the field related to scene text recognition and the strength of AI technology in Taiwan.


2.Content of the competition

The competition is held from the end of July to mid-October. The challenge of the competition includes the multi-type Chinese characters, tilted signboard, different sizes of Chinese characters, object shielding, text-like pattern texture interference, light and shadows. In addition, compared with a small amount of alphanumeric characters, the number of traditional Chinese characters appearing in the street may be in the thousands, which will increase the difficulty of text recognition. We expect that the contestants use the deep learning/image recognition instead of dictionary corrections. Welcome to join us!


3.Registration time:2021/07/28-2021/09/30


4.Period of competition:2021/08/04-2021/10/07



6.Bonus:

The first place of the student group:$40000

The second place of the student group:$30000

The third place of the student group:$20000

The Excellence Award of the student group (x3):$10000

The Honorable Mention Award of the student group (x6):$5000

趨勢科技人工智慧菁英獎:$30000

趨勢科技創意獎:$30000

A team with a ranking in the top 25% and a score exceeding the baseline will be awarded a certificate of MOE.

unnamed (3).jpg
TBrain_logo_960x200.png

Cooperation Platform

​Artificial Intelligence Co-creation Platform AIdea

​Cooperation Platform

T-Brain AI in action

Ministry of Education Artificial Intelligence Competition and Annotation Data Collection Project Office

Phone: (03)422-7151 #35323
E-mail: moe.ai.ncu@gmail.com

©2018-2021 AI CUP
  • Instagram
  • Facebook
bottom of page