Yuan Yao 姚 远

Hi! I am a second-year Ph. D. student at University of British Columbia, working on Computer Graphics and Computer Vision. Welcome to my academic webpage!

Email  /  CV  /  Google Scholar  /  LinkedIn

I entered into the world of Computer Graphics when I was just 3 years old by playing the game "Rayman". I also spent more than 10 years on Mathematics Olympiad and got many prizes. Therefore, Computer Science and Math are both my favorite subjects. Besides, I also love music, I can play guitar, piano, clarinet and sing. I wish I can finally bring techniques to the arts.

I got my bachelor degree in computer science from Shanghai Jiao Tong University, during which I spent two months studying in University of California, Los Angeles and three months doing research in University of Illinois Urbana-Champaign. Besides, I also fortunately interned at Unity Technologies, Microsoft Research Asia and Megvii Research. You can check my CV for more details.

My research interest lies on the intersection of Computer Graphics and Computer Vision. More specifically, I am now actively working on 3D Vision, Geometry Processing and Physics-based Simulation. Previously I worked on Computational Photography and Virtual Reality. I always have many interesting ideas in different topics, feel free to check out my works or reach out for collaboration(even just a discussion).

I am co-supervised by Prof. Alla Sheffer and Prof. Leonid Sigal, and also working closely with Prof. Helge Rhodin.

My research is founded by NSERC and Vector Institute.

An explicit single-view 3D reconstruction method by estimating back view 2.5D maps from front view.

A non-parametric algorithm for neural style transfer.

A virtual gestural interface developed using Leap Motion.

Here are some course projects during my graduate study, I also did a lot of fun projects in my undergrad, check here.

An isotropic remeshing algorithm which allows adaptive edge length and preserve features. The codes were written based on libigl.

We proposed a two-step music-aligned dance sequence generation method.

Corotational Finite Element Method based Cloth Simulation using Linearly-Implicit time integration.

A simple 3D face reconstruction method from single view image. We use the parametrized model "FLAME" and minimize the distance between 2D landmarks and projected 3D landmarks.

Here are a few other open source codes I wrote during learning/researching.

This webpage template credits to Jon Barron