I'm Ming-Yu Liu, Vice President of Research at NVIDIA and an IEEE Fellow. I lead the Deep Imagination Research group and the NVIDIA Cosmos effort, where we advance Generative AI for Physical AI — building world foundation models that allow machines not just to perceive the world, but to simulate, reason about, and interact with it.
Over the years, my team has helped define multiple new product categories, from generative models powering global creative ecosystems to real-time neural graphics systems used by millions. We publish regularly at the top AI and graphics venues, but more importantly, we translate research into platforms that developers and enterprises can build on.
At a high level, I'm driven by the vision of building the "Matrix" for robots — a physics-grounded simulated universe where machines can dream, rehearse, and accumulate experience safely at scale before ever acting in the real world.
Awards
- Best of CES, Best of AI, CNET 2025
- Best Inventions, TIME 2023
- Distinguished Alumni Award, University of Maryland College Park, ECE, 2023
- Best in Show Award, SIGGRAPH Real-Time Live 2021
- Most Disruptive Inventor Award, Forbes 2020
- Best of What's New Award, Popular Science 2019
- Best in Show & Audience Choice Award, SIGGRAPH Real-Time Live 2019
- Best Paper Finalist, CVPR 2019
- 1st place, Domain Adaptation for Semantic Segmentation, WAD Challenge, CVPR 2018
- 1st place, Optical Flow Competition, Robust Vision Challenge, CVPR 2018
- Outstanding Reviewer, CVPR 2018
- Pioneer Research Award, NVIDIA 2018
- Pioneer Research Award, NVIDIA 2017
- NTECH Best Presenter Award, NVIDIA 2017
- CR&D Award, Mitsubishi Electric Research Labs (MERL) 2016
- Best Paper Finalist, Robotics Science and Systems (RSS) 2015
- R&D 100 Award, R&D Magazine 2014