By: Jin Song, Peak Associate

SIGGRAPH (Special Interest Group on Computer Graphics and Interactive Techniques) is the largest computer graphics conference in the world, bringing together scientists, artists, and other professionals working on cutting-edge technologies in the field. It has hosted guest speakers like Mark Zuckerberg and Jensen Huang. This year, it was held in August in Vancouver.

I learned about the conference’s existence and the fact that it would be local in CMPT 361 (Introduction to Visual Computing), which I took last year. I heard about it once more in CMPT 466 (Computer Animation), and my professors — who were also in attendance — encouraged the class to apply to be student volunteers if we were interested. 

Indeed, from a young age, I had been interested in computer graphics (specifically, animation), so I was naturally intrigued. I applied to be a student volunteer and was accepted. It was one of the coolest weeks of my life, and I interviewed three other SFU students about their experiences with the conference.

Michael Xu, a computing science PhD student, gave a presentation on his contribution, PARC: Physics-based Augmentation with Reinforcement Learning for Character Controllers. He said he was motivated by his desire to “have a physically simulated character that can traverse a Minecraft world,” which was not present in past research. PARC is a physics-based method for iteratively training an animation machine learning model. It uses the model’s output as the input for the next cycle, progressively improving and expanding the motion dataset. 

My other two interviewees were my classmates from my computer graphics courses, Tyrus Tracey and Peter Soava. Tracey — who presented a poster at SIGGRAPH — is a fourth-year computing science student like myself, while Soava, a fourth-year software systems major, was another student volunteer.

Tracey described his contribution to the conference, titled Physically-Based Compositing of 2D Graphics, as follows: “It is basically a new type of image editing technique. Given an image, you can insert a new image onto that image. It will conform to, say, the contours of the table or the wall or lighting present, that sort of thing.” He and his classmate built a tool on top of prior work to “process a bit more usable for average people.”

Attending Xu’s presentation, looking at Tracey’s poster, and seeing all of the other SFU contributors at SIGGRAPH 2025, I was hit by a profound sense of awe. On one hand, I was incredibly impressed by the work, on a technical level. These cutting-edge innovations were truly spectacular. On the other hand, I felt proud to be part of the SFU community, and was astonished to see its influence. Having grown up in Vancouver, I had always considered it familiar and homey, so seeing it in this new, influential light was pleasantly surprising to me. 

Soava said, “I think it was really cool bumping into a bunch of industry people. It seems kind of unique because Vancouver is such a VFX hub, like you have ILM, Sony Image Works, Coalition, CD Projekt Red, EA, also headquartered here. It’s just like, you asked somebody who has worked on all these big projects, and it turns out that they were local.”

When asked if they would recommend attending, Soava said, “If you find the stuff even mildly cool, you just spend a week with a bunch of extremely cool people just talking about extremely cool stuff.” Xu said

“If you’re just interested in animation, graphics, and art, SIGGRAPH is a good place because it’s more than just a technical conference.”

— Michael Xu, SFU Computing Science PhD Student

SIGGRAPH 2026 will be held in Los Angeles from July 19 to 23.