In the final of the Global AI Transmission Competition (AI-Trans) held in Beijing on January 12, Kuyusi and WwW, two teams from CUHK-Shenzhen, with their professionalism in the design of algorithms and outstanding performance in the presentation, have been highly appreciated by all judges, winning the “Best Non-Machine Learning Award” and the “Best Demonstration Award” respectively. In addition, Mr. Li Yupeng, instructor of the teams and lecturer from Shenzhen Research Institute of Big Data ((SRIBD) in CUHK-Shenzhen was awarded “The Best Instructor”.

AI-Trans is a competition led by the Department of Computer Science and Technology of Tsinghua University, co-organized by the The Computer Society of the Republic Of China (CSROC), the Intelligent Information Network Committee of The Chinese Association for Artificial Intelligence (CAAI), the Information Communication Network Committee of China Institute of Communications (CIC) and Power Info and with the participation of The Stanford Computer Science Department, NUS Computing and Future Network Theory Lab of Huawei as partners.

The competetion attracted hundreds of teams from all over the world, including teams from top universities in China, including Chinese Academy of Sciences, Tsinghua University, Peking University, Shanghai Jiaotong University, Beijing University of Posts and Telecommunications, Tianjin University, Wuhan University, and professional teams from the industry such as Kuaishou and Alibaba. There were also participants from the United States, Canada, Germany, Japan participating.

Three teams from CUHK-Shenzhen Kuyusi (Li Yijing, Shi Siyuan, Wang Yiran), WwW (Hao Senyue, Xie Guochao, Mo Ziheng) and Taiyangshen (He Tao, Li Xiaoxuan, Ran Chenyu) led by Mr. Li YuPeng, lecturer from Shenzhen Research Institute of Big Data ((SRIBD) in CUHK-Shenzhen joined the competition. After more than two months of meticulous preparation, the three teams performed well in the preliminary and all advanced to the final along with other 22 teams.

Against new challenges presented by the explosive growth of data scale and the personalized demand of dynamic heterogeneous networks, network researchers, equipment vendors and Internet service providers have been trying to explore a variety of emerging technologies to advance intelligent network transport, ushering in new opportunities for "intelligent" network transmission technologies driven by recent technological innovations. For example, artificial intelligence is becoming the mainstream research direction for the new industrial revolution. In the era of big data, Shenzhen Research Institute of Big Data (SRIBD) in CUHK-Shenzhen is dedicated to achieving automated, efficient, accurate and fair intelligent network transmission through machine learning, data mining, and cybernetics.

Currently, domestic popular live broadcast platforms such as Douyu and Yingke provide default choices for higher code rates. When there are frequent lags, a corresponding pop-up window will appear to prompt users to downgrade to SD or smooth mode. There are also platforms providing fixed low code rate, so users cannot enjoy a high-definition viewing experience. Academics and industry have been actively studying adaptive adjustment rate algorithm in recent years. AI-Trans has set up the corresponding platform, welcoming participants to utilize the power of technology to solve real-life problems.

From left：Li Yupeng（Instructor），Hao Senyue，Shi Siyuan，Xie Guochao，Mo Ziheng