iQIYI Announces Winners of its iCartoonFace Challenge, the First Large-scale Cartoon Character Recognition Competition

World-leading iCartoonFace datasets drive intelligent recognition technology

iQIYI Inc., an innovative market-leading online entertainment service in China, today announced the winners of its iCartoonFace Challenge (the “Challenge”), the first large-scale cartoon character recognition competition in China. The team from Southeast University won the Challenge with a MAP (Mean Average Precision) of 0.9291 in the cartoon character detection track and the teams from Zhejiang University and Sun Yat-sen University jointly claimed the title of Best Algorithm in the cartoon character recognition track with a Rank1 accuracy of 92.4697%.

The Challenge, held in partnership with the 29th International Joint Conference on Artificial Intelligence (IJCAI) and the 17th Pacific Rim International Conference on Artificial Intelligence (PRICAI), was held over a period of two months, attracting a total of 481 participating teams from top universities and leading technology firms in China. Contestants included Peking University, Fudan University, Zhejiang University, Shanghai Jiaotong University, University of Chinese Academy of Sciences, as well as renowned companies such as Alibaba, Tencent, Baidu, JD.com and SenseTime.

Recommended AI News: Comscore Research Shows Increased Consumer Interest In Online Vehicle Shopping And At-Home Delivery As Industry Continues To Recover

While human facial recognition technology is becoming increasingly sophisticated and mature, cartoon character recognition is still in its infancy. Cartoon characters are far more complicated and diverse than human faces in terms of both color and texture. Other challenges include the fact that cartoon characters from the same content or universe could be highly similar in appearance, but characters from different universes may vastly contrast. Due to the complexity of this dataset, cartoon character recognition poses far more challenges than human facial recognition.

Currently, most of the existing cartoon datasets for recognition suffer from issues such as lack of data volume or noisy data. During the Challenge, iQIYI offered contestants access to the largest known manually tagged datasets in the world for cartoon detection and recognition that contains over 5,000 cartoon characters and more than 400,000 high-quality and authentic images.

Recommended AI News: Unbound Tech Announces Key Security-Rooted Hires To Support North American Market Growth

iQIYI is committed to researching cutting-edge technologies and exploring their applications in entertainment. iQIYI has in the past two years successfully held two video-based multimodal biometric recognition challenges which helped significantly improve the accuracy of multimodal recognition technology. Currently, cartoon character recognition has been applied to multiple innovative businesses and features of iQIYI’s. For example, AIWorks can automatically collect cartoon character content to generate a mixed-cut video based on a specific theme; and a feature known as Qiguan (AI Radar) allows users to identify the cartoon characters in an animated show or movie.

Through the iCartoonFace Challenge held in partnership with IJCAI and PRICAI, iQIYI has achieved further breakthroughs in the field of cartoon detection and recognition, promoted the development of cartoon character recognition technology, enabled a deep integration of technology and content that could lead to more usage scenarios, thereby advancing the ongoing evolution of intelligent detection and recognition technology.

Recommended AI News: CloudBees Named A Leader For Continuous Delivery Release Automation By Independent Research Firm