[학술/연구] Physical AI 석학 초청 세미나
- 신공학관 161호
- 2026.07.10 ~ 2026.07.10 15:00 ~ 16:20
- 98
Physical AI 분야의 석학 교수인 미국 매릴랜드 대학의 Dinesh Manocha 교수님과 Ming Lin 교수님을 모시고 특강을 진행하려고 하니, 관심 있는 분들의 많은 참여를 바랍니다.
장소: 신공학관 161호
연락처: 김영준 교수 (kimy@ewha.ac.kr)
Robot Navigation in the Wild
Dinesh Manocha
University of Maryland at College Park
https://gamma.umd.edu
Abstract:
In the last few decades, most robotics success stories have been limited to structured or controlled environments. A major challenge is to develop robot systems that can operate in complex or unstructured environments corresponding to homes, dense traffic, outdoor terrains, public places, etc. In this talk, we give an overview of our ongoing work on developing robust planning and navigation technologies that use recent advances in computer vision, sensor technologies, machine learning, and motion planning algorithms. We present new methods that utilize multi-modal observations from an RGB camera, 3D LiDAR, and robot odometry for scene perception, along with deep reinforcement learning for reliable planning. The latter is also used to compute dynamically feasible and spatial aware velocities for a robot navigating among mobile obstacles and uneven terrains. We have integrated these methods with wheeled robots, home robots, and legged platforms and highlight their performance in crowded indoor scenes, home environments, and dense outdoor terrains. We will highlight the benefits for social navigation.
BIO: Prof. Dinesh Manocha is Paul Chrisman-Iribe Chair in Computer Science & ECE and Distinguished University Professor at University of Maryland College Park. His research interests include virtual environments, physics-based modeling, and robotics. His group has developed several software packages that are standard and licensed to 60+ commercial vendors. He has published more than 950 papers & supervised 67 PhD dissertations. He is a Fellow of AAAI, AAAS, ACM, IEEE, and NAI and member of ACM SIGGRAPH and IEEE VR Academies, and Bézier Award from Solid Modeling Association. He received the Distinguished Alumni Award from IIT Delhi the Distinguished Career in Computer Science Award from Washington Academy of Sciences. He was a co-founder of Impulsonic, a developer of physics-based audio simulation technologies, which was acquired by Valve Inc in November of 2016.
Dynamics-Informed Learning in the Physical World
Ming C. Lin
University of Maryland at College Park & Amazon FAR
http://www.cs.umd.edu/~lin/
http://gamma.umd.edu/
ABSTRACT: Recent progress in data-driven revolution has led to significant advances in artificial intelligence (AI), while many challenges remain in terms of computational efficiency, AI safety, and generalization. In this talk, we present latest advances to address some of these issues. First, we introduce efficient Time-Aware World Model (TAWM) that minimizes training costs with improved accuracy by adaptively sampling multiple time steps across scale for greener computing and energy efficiency for generalist robots and visual odometry of mobile robots. Next, we introduce the first end-to-end framework that integrates linear-temporal logic with differentiable simulation, enabling efficient gradient-based learning directly from formal specification. We also present an immersive vehicle-traffic coupled simulation system that models diverse sets of adversarial scenarios, including accidents, extreme weather conditions, human behaviors, and unexpected events, to sample and capture driving data in tail-end distribution. Lastly, we discuss techniques to construct a generalizable 3D foundation model of garments for virtual try-on. Together these methods showcase a collection of data generation and sampling strategies that can provide more computational efficient, representative, and reliable data synthesis for robust and generalizable learning. They offer new insights for addressing some data challenges for more robust and effective learning from simulation and observations. I conclude by discussing some possible future directions and data challenges for learning in the Physical World.
SHORT BIOGRAPHY: Ming C. Lin is currently Distinguished University Professor, Dr. Barry Mersky and Capital One E-Nnovate Endowed Professor, former Elizabeth Stevinson Iribe Chair of Computer Science at the University of Maryland College Park, and John R. & Louise S. Parker Distinguished Professor Emerita of Computer Science at the University of North Carolina (UNC), Chapel Hill. She is also an Amazon Scholar. She obtained her B.S., M.S., and Ph.D. in Electrical Engineering and Computer Science from the University of California, Berkeley. She received several honors and awards, including the NSF Young Faculty Career Award in 1995, Honda Research Initiation Award in 1997, UNC/IBM Junior Faculty Development Award in 1999, UNC Hettleman Award for Scholarly Achievements in 2003, Beverly W. Long Distinguished Professorship 2007-2010, UNC WOWS Scholar 2009-2011, IEEE VGTC Virtual Reality Technical Achievement Award in 2010, Washington Academy of Science Distinguished Career Award 2020, ACM SIGGRAPH Seminal Graphics Paper in Physical Simulation, IEEE ICRA 2026 Most Influential Paper Award, and many best paper awards. She is a Fellow of National Academy of Inventors, ACM, IEEE, Eurographics, ACM SIGGRAPH Academy, and IEEE VR Academy.