ZhihuAbout Me
I am Chunyang (Cedric) Zhang, a Ph.D. candidate in Computer Science at the University of New South Wales (UNSW), Canberra, advised by Prof. Daoyi Dong and Prof. Huadong Mo. My research lies at the intersection of Generative AI and Optimization Theory.
My long-term goal is to build embodied AI systems that act reliably in the physical world. I focus on Large Multi-Modal Models that combine the perceptual and reasoning capacity of generative models with the decision-making rigor of optimization and control. The next leap in real-world deployment, I believe, will come from systems that not only generate plausible behavior, but also optimize it under physical, safety, and resource constraints.
Education
-
Ph.D. in Computer Science
-
M.Eng. in Control Science & Engineering
-
B.S. in Weapon Science & Technology
Research Interests
- Multi-Modal Foundation Models: Vision–Language–Action (VLA) models, world models, and embodied policy learning for real-world robotics and autonomous systems.
- Generative AI: Diffusion models, video generation, and consistent multi-instance synthesis — the perception and simulation engine that feeds downstream decision making.
- Optimization & Control: Reinforcement learning, distributed optimization, and intelligent control — the decision-making backbone that turns generative perception into safe, executable actions.
- AI for Science: Physics-informed machine learning and neural operators bridging first-principle dynamics with data-driven models for stronger out-of-distribution generalization.
I am actively looking for collaborations that push generative intelligence into closed-loop, physically grounded systems — particularly in robotics, autonomous driving, and large-scale industrial control.
News
-
Joined the School of Systems and Computing at UNSW as a Ph.D. candidate.
-
Our paper “Robust Control of Multi-Line Re-Entrant Manufacturing Plants...” was published in IEEE Transactions on Automation Science and Engineering (T-ASE).