Develop AI Talents for the Future

AI Challenger Global AI Contest

“AI Challenger” is a platform initiated by Sinovation Ventures and supported by strong industry and academic resources, for open datasets and programming competitions to artificial intelligence (AI) talents around the world. The goal of AI Challenger is to empower worldwide AI researcher, scientist, and developer talents to advance their AI research and technological development without the constraints of data resources. In the first year, AI Challenger 2017 released over 1 million data entries based on 300,000 images, as well as over 10 millions language-based data. 8892 teams from 65 countries competed and the winners shared over 300,000 USD prize. All these made AI Challenger the largest dataset platform for scientific research and the largest non-commercial competition platform in China.

AI Challenger 2018

Working with more leading AI companies, universities, and government organizations, AI Challenger 2018 will offer new datasets in the scientific research and industry application frontiers:

Main datasets and competitions:
  • Autonomous driving
  • Reasoning questions and answers
  • Machine Learning for local services
  • 2017 datasets’ upgrade
Experimental datasets and competitions:

zero-shot learning, robotics controlling, weather forecast, quantitative trading and etc.

AI Challenger 2018 main datasets and competitions will kick-off in August and last till December. Experimental datasets and competitions are running through the year.

  • August 2018, main datasets and competitions kick-off.
  • August to late September 2018: Training dataset and validation dataset is available for download.
  • Late September to mid-November 2018: Test dataset A will be available for download. Until the test dataset B is released, all scores on leaderboard will be calculated on submission result of test dataset A.
  • Mid-November 2018: Test dataset B will be released. The final leaderboard will be announced after scoring is completed.
  • Early December 2018: finals and award ceremony.
  • Experimental datasets and competitions open every quater.
  • Online and offline tech talks every month.
  • Online community launch in May 2018.

As the online community and practice playground, AI Challenger also closely supports The International AI Training Program For Chinese Universities, which plans to directly train over 500 school faculties and 5,000 students, and impact hundreds of thousands more AI talents in 5 years.

AI Challenger 2017

Launching for the first time in fall 2017, AI Challenger is chartered and organized by Sinovation Ventures, Sogou, and Toutiao (ByetDance), three heavy weight players headquartered in Beijing from venture capital, search, and media technology domain expertise respectively. AI Challenger 2017 released over 10 millions language-based data, over 1 million data entries based on 300,000 images, as well as over US $300,000 prize. It attracted one million website visits from 120 countries, and 8892 teams from 65 countries participated the competitions, which made AI Challenger the largest dataset platform for scientific research and the largest non-commercial competition platform in China. 

AI Challenger 2017 covered about 347 universities (247 domestic and 100 overseas), including Tsinghua University, Peking University, Chinese Acedamy of Sciences, Shanghai Jiao Tong University, Fudan University, Hong Kong University of Science and Technology, National Taiwan University, Massachusetts Institute of Technology, Stanford University, Carnegie Mellon University, Cornell University, Imperial College London, University of Cambridge, University of Toronto, University of Tokyo, University of Sydney. And 273 companies got covered, including Baidu, Tencent, Alibaba, iFlytek,, Meituan, Didi, Xiaomi, Cheetah Mobile, UISEE, Mobike, Microsoft, Amazon, GE, Intel, Samsung, Cisco and etc.

AI Challenger 2017 was co-chaired by Sinovation Ventures CEO Kai-Fu Lee, Sogou CEO Xiaochuan Wang, Toutiao CEO Yiming Zhang. Judges committee included UC Berkeley Professor and IEEE Fellow Yi Ma, Peking University Tenure Professor and winner of National Science Fund for Distingushed Young Scholars Yizhou Wang, Founder and ACL Fellow Dekang Lin, Megvii Chief Scientist Jian Sun, and other academic and industry experts, offering comments and advice to the contestants.

AI Challenger 2017 set five main competition tracks: Human Skeletal System Keypoints Detection, Scene Classification, Image Captioning (Chinese), English-Chinese Machine Translation, English-Chinese Simultaneous Interpretation; and one experimental competition track, Virtual Stock Trend Forecast. These competitions were designed by top AI experts, and then proessed by the organizers with in-depth industry research, which ensured the meanings for real-world problem solving.

AI Challenger 2017 Promo Video AI Challenger 2017 Winner List