Huu Dat Do

I am an undergraduate student at VinUniversity, working closely with Professor Wray Buntine and Professor Laurent El Ghaoui. I had the opportunity to intern at the Advanced Machine Intelligence Lab (AMI Lab) at KAIST under the supervision of Professor Tae Hyun Oh and collaborated with Professor Minsu Cho from POSTECH. Additionally, I am conducting research with Professor Luu Anh Tuan at NTU.

My research focuses on understanding and reconstructing intelligence through computational methods. Specifically:

Compositional understanding: Humans understand the world with different modalities (language, vision, sound, touch, etc) in a compositional, synchronous manner. Can we build AI systems that can learn compositional, generalizable representations across different modal domains and perform many novel and complex tasks?
Compositional generation: Imagination and reasoning play a vital role in human-level intelligence. How can we develop AI systems that faithfully generate novel data in “imagination” and use it as mental simulation engines supporting understanding, learning, reasoning, and even complex generation itself?
Efficient ML: The human brain operates on an extremely low power supply and learns from data much more efficiently. Can we incorporate other natural priors such as symmetries, equivariance, and iterative inference into the AI system to learn more efficiently and still be scalable?

I am also passionate about applying AI to tackle global challenges in robotics, physics, economics, climate change, and medical science.

selected publications