Dr. Zhengyou Zhang
Modular Embodied Intelligence
We draw inspiration from Kahneman’s human intelligence theory about system 1 and system 2 thinking, as well as from the knowledge of human brain, and propose a modular architecture for embodied intelligence and a hierarchical learning framework, called “SLAP”, to enable robots to exhibit lifelike agility and strategy in complex environments. Our approach moves beyond conventional task-specific controllers and end-to-end RL methods by pre-training generative models on animal/human motion datasets, thus retaining extensive knowledge of animal/human behavior. The framework goes from this environment-agnostic primitive-level controller, up through the environment-level controller to adapt to various environments, and finally into the strategic-level controller to tackle complex downstream tasks. This flexible framework allows for continuous accumulation of knowledge at different levels without affecting the usage of other knowledge levels. We successfully apply the multi-level controllers to the quadrupedal robots and bipedal humanoids that navigate complex obstacles and accomplish complex tasks, where lifelike agility and strategies emerge. This research advances the field of robot control by offering new insights into the reuse of multi-level pre-trained knowledge and the effective tackling of complex real-world tasks.
Bio: Dr. Zhengyou Zhang (ACM Fellow and IEEE Fellow) is the Chief Scientist at Tencent, China, and the Director of Tencent Robotics X since March 2018. Before that, he was a Partner Research Manager with Microsoft Research, Redmond, WA, USA, for 20 years. Before joining Microsoft Research in March 1998, he was a Senior Research Scientist with INRIA (French National Institute for Research in Computer Science and Control), France. In 1996-1997, he spent a one-year sabbatical as an Invited Researcher with the Advanced Telecommunications Research Institute International (ATR), Kyoto, Japan. He is the Founding Editor-in-Chief of the IEEE Transactions on Cognitive and Developmental Systems, is on the Honorary Board of the International Journal of Computer Vision and on the Steering Committee of the Machine Vision and Applications, and serves or served as an Associate Editor for many journals. He received the IEEE Helmholtz Test of Time Award for his paper published in 1999 on camera calibration, now known as Zhang’s method, and the AISTATS Test of Time Award for his paper published in 2015 on Deeply-Supervised Nets (DSN). According to Google Scholar, his h-index is 108, and he has over 78K citations. He has over 250 patents issued.
Prof. Kenji Hashimoto
Humanoid Robots in Action: From Research Bench to Real-World Impact
This keynote will open by showcasing the expansive impact of humanoid robot research through a diverse array of real-world applications. I will explore how advancements in this field have significantly contributed to sports science, psychology, disaster response, and entertainment, demonstrating these complex systems’ broad utility and evolving capabilities. This exploration of practical applications naturally leads to cutting-edge research at our laboratory, Waseda University. Our central aim is to realize truly agile robots that can seamlessly adapt to complex and dynamic environments, exhibiting remarkable physical dexterity and the intellectual capacity to address challenging tasks effectively. To this end, I will detail our current innovations, including the development of advanced locomotion systems exemplified by the multi-modal four-limbed robot WAREC-1 for disaster response, capable of diverse movements like quadrupedal, bipedal, crawling, and ladder climbing. Further, our work extends to pioneering jumping-legged robots and versatile hybrid leg-wheel designs. Crucially, I will also highlight our contributions to artificial intelligence for sophisticated environmental perception, which is fundamental for achieving the autonomy of these mobile systems. This presentation will thus demonstrate how our integrated approach – from understanding real-world needs to developing groundbreaking robotic solutions – is pushing the boundaries of mobile robotics for profound societal impact.
Bio: Kenji Hashimoto received his B.E., M.E., and Ph.D. degrees from Waseda University, Japan, in 2004, 2006, and 2009, respectively. His international experience includes a Postdoctoral Researcher position at College de France-CNRS, France, from 2012 to 2013. He subsequently held academic appointments at Waseda Institute for Advanced Study, Waseda University (Assistant and Associate Professor, 2015-2018), and Meiji University (Associate Professor, 2018-2022). Since September 2022, he has been a Professor at the Graduate School of Information, Production and Systems (IPS), Waseda University, where his research focuses on advanced legged and humanoid robots. Prof. Hashimoto’s pioneering work in these fields has been recognized with numerous accolades, including the IEEE Robotics and Automation Society Japan Chapter Young Award (2006), the JSME Fellow Award for Outstanding Young Engineers (2008), the RSJ Young Investigation Excellence Award (2015), and the Kisoi Motohiro Award for Academic Achievement (2018). He is an active member of various robotics-related academic societies such as IEEE, IFToMM, RSJ, and JSME.