After a brief pivot to online conferencing (convenient, but not nearly as fun), SIGGRAPH made its triumphant return in person, hosted by the wonderful city of Vancouver. An absolute must on the industry calendar for celebrating advances in computer graphics and interactive techniques, it was once again the nexus for thousands of passionate innovators to meet, share ideas, and learn from each other.
Not only did we reconnect with old friends and discover new technologies, we introduced our latest innovations to an evolving creator community. Over the course of the week, we unveiled some of our work through a series of talks, courses, technical papers, and demos at our booth. Colleagues joined together from across Unity, globally, to discuss everything from technical updates and hyperrealistic digital humans to advances in real-time rendering for games, solutions for virtual production, VFX, and so much more.
With a large number of presentations covering topics from facial autorigging for stylized characters and realistic digital humans to VFX for foliage and vegetation, we decided to compile some of our favorite moments.
Now, let’s relive the action with our recap.
We premiered our latest RT3D demo, Lion, at SIGGRAPH 2022. A joint project between Unity and the Monster Emporium Animation School, Lion features innovations in real-time technology through content created collaboratively for the first time across Wētā Digital, Ziva, SyncSketch, SpeedTree, and the Unity Graphics team.
Lion was born from our research and development, alongside our growing ecosystem of real-time artistry tools and features. It underscores what’s possible when all these tools work together.
You can read this recent blog to learn more about the technologies used in the Lion demo.
Having your session accepted into Real-Time Live! is always an honor, let alone winning one of their awards. But that’s exactly what happened to the Unity Labs ML Artistry team. Presented by Senior Machine Learning Developer Florent Bocquelet, “AI and physics assisted character pose authoring” won the audience vote for new research on character animation and posing through machine learning.
The presentation showcased a tool that enables Unity users with minimal artistic experience to leverage both machine learning and a physics engine to author natural-looking poses in seconds. We’re very proud of the team for this accomplishment, and want to give a shoutout to all the other speakers for their fascinating presentations as well.
Our demo Enemies underwent further refinement leading up to the conference. Thanks to the addition of Ziva’s powerful facial animation software and a new, strand-based Hair Solution (now available to download via GitHub as an experimental package), it was ready to go.
Selected as a featured project part of the prestigious Real-Time Live! and Electronic Theater programs, we demonstrated the real-time animation workflow for Enemies, as well as Unity’s latest advances for creating digital humans. In the live session, Unity’s Krasimir Nechevski and Daniel Gagiu exhibited the quick editing process behind Enemies’ protagonist, Louise – and more specifically, her facial performance in Maya with Ziva’s control scheme to instantly deploy the new performance in Unity. They then switched over from the protagonist to the new character, Emma, emphasizing the extent to which Ziva can address key data.
Check out the making of Enemies.
SIGGRAPH also saw the launch of Ziva Face Trainer (ZFT), one of our latest tools for rapidly creating 3D human and character facial animations. During a deep dive into the newly released autorigging solution for RT3D faces, we highlighted the benefits of ZFT’s deep-learned parameter space, ARKit, and expression controls.
Unity Virtual Production Supervisor Habib Zargarpour discussed how he planned specific shots for his film projects, Dark Asset and Comandante, both made with virtual production tools across the Unity platform. Alongside VFX Artist Setareh Samandari, we were given a behind-the-scenes look at multiple RT3D productions, including the previs-to-layout approach used for the short film Leap and Kanye West’s 24 music video.
The three-part “Advances in real-time rendering” course, organized by Natalya Tatarchuk, Unity Distinguished Technical Fellow and Chief Architect, VP, Professional Artistry and Graphics Innovation, has now been running for 17 years at SIGGRAPH. Covering production-proven techniques for the interactive rendering of complex and engaging virtual worlds in video games, it remains a fan favorite today.
During a fascinating technical paper review, the Wētā Digital team introduced Loki, a new framework for the robust simulation of fluid, rigid, and deformable objects. Their second paper then focused on a physically-based simulation method for bubbles and foam with whitewater effects. This makes the bubbles appear more naturally connected to the underlying water without the need for excessive expert parameter tuning by artists.
Those who visited Unity’s booth on the show floor got to try some of our artist tools and workflows firsthand, exploring the latest solutions for efficient creation and collaboration. Through interactive demos, visitors experienced:
Finally, we want to thank everyone who attended the Unity sessions and visited our booth this year. We already can’t wait to see you at the next SIGGRAPH.