Dongbin Zhao

Also known as 赵冬斌

Dongbin Zhao is a reinforcement-learning researcher at the Chinese Academy of Sciences whose group applies deep and multi-agent reinforcement learning to games, driving and robots. He earned degrees at Harbin Institute of Technology in 1994, 1996 and 2000, was a postdoctoral fellow at Tsinghua University from 2000 to 2002, and joined the Institute of Automation in 2002. A survey he co-authored on deep reinforcement learning and computer Go won a 2018 outstanding paper award from the journal Control Theory and Applications, and in 2019 he co-authored a StarCraft micromanagement method published in IEEE Transactions on Emerging Topics in Computational Intelligence. In 2020, he became an IEEE Fellow, and in 2022 he was General Chair of the IEEE Conference on Games in Beijing. Today, he is a researcher at the Institute of Automation and a professor at the University of Chinese Academy of Sciences, where his work covers deep and multi-agent reinforcement learning, large models, embodied intelligence and smart driving.

Now at: University of Chinese Academy of Sciences and Institute of Automation, Chinese Academy of Sciences
Badges: Loyal

Timeline

1990–2002

Harbin degrees and a Tsinghua postdoc

He earns bachelor's, master's and doctoral degrees at Harbin Institute of Technology in 1994, 1996 and 2000. From 2000 to 2002, he is a postdoctoral fellow at Tsinghua University.
2002–2012

Joining the Institute of Automation

He joins the Institute of Automation of the Chinese Academy of Sciences in April 2002 as an associate researcher. From 2007 to 2008, he is a visiting scholar at the University of Arizona, and in November 2012 he becomes a full researcher and doctoral supervisor at the institute.
2018

Survey of deep reinforcement learning

In 2018, a survey of deep reinforcement learning and the development of computer Go, which he co-authors with Kun Shao, Yuanheng Zhu and others, is one of two papers to win the outstanding paper award of the journal Control Theory and Applications.
2019

StarCraft micromanagement by reinforcement learning

In 2019, he co-authors with Kun Shao and Yuanheng Zhu a StarCraft micromanagement method that combines reinforcement learning and curriculum transfer learning, published in IEEE Transactions on Emerging Topics in Computational Intelligence. In small scenarios, the trained units defeat the built-in game AI with 100% win rates, and DeepMind cites the paper in its Nature article on AlphaStar.
2020

Becoming an IEEE Fellow

He becomes an IEEE Fellow in 2020. His group also wins three championships at the 2020 RoboMaster AI Challenge.
2022

Beijing prize and Games chair

In 2022, he ranks first among the recipients of a Beijing Science and Technology Award, second prize in natural science, for efficient deep reinforcement learning algorithms and their optimality analysis. He is General Chair of the IEEE Conference on Games held in Beijing in August 2022. IEEE Transactions on Emerging Topics in Computational Intelligence gives the StarCraft micromanagement paper its 2022 outstanding paper award.
2025–2026

BAAI Scholar and IEEE lectures

In 2025, he is named a BAAI Scholar by the Beijing Academy of Artificial Intelligence and receives the Chinese Academy of Sciences Li Pei Outstanding Teacher Award. In 2026, he is part of the IEEE Computational Intelligence Society Distinguished Lecture Program.
As of October 2026

Professor at CASIA and UCAS

He is a researcher at the Key Laboratory of Multimodal Artificial Intelligence Systems of the Institute of Automation, Chinese Academy of Sciences, and a professor at the University of Chinese Academy of Sciences. His work covers deep and multi-agent reinforcement learning, large models and embodied intelligence, smart driving, game AI and robotics. He is also an associate editor of IEEE Transactions on Systems, Man, and Cybernetics: Systems and IEEE Transactions on Artificial Intelligence.