|
Martin Guay
|
My background is modeling / AI ML, graphics, vision & simulation.
After studying math, I started my career building video games software in Montreal.
I realized I wanted to work on the new stuff and did a Master’s on simulating fluids (fire and liquids).
Then I moved to France to pursue a PhD on 3D character animation at INRIA. I then joined Disney Research in Zurich where I built simulated character control using deep reinforcement learning, new augmented reality concepts such as virtual try-on,
and finally a live AI talking avatar guests can speak with, which ultimately was purchased by
NVIDIA.
In 2017, I saw the potential of deep learning and jumped into computer vision for tracking actors in camera for VFX.
Today I lead R&D efforts in animation and VFX at
Disney Research . Our team is using deep learning to built a full stack best-in-class actor tracking systems for VFX, and motion completion for animation; both of which are now used by ILM and Pixar respectively.
List of Publications
Augmenting Cats and
Dogs
Computer Vision, Imaging and Computer Graphics Theory and Applications
(VISIGRAPP), 2021.
D. Borer, N. Isik, J. Buhmann, M. Guay
Authoring Motion Cycles
Symposium on Computer Animation (SCA), 2017.
Loïc Ciccone, Martin Guay, Mauri Nitti, Robert W. Sumner