SIGGRAPH 2022—one of the world’s premier conferences on computer graphics and interactive technologies—concluded its state-of-the-art conference on August 11, 2022.
The event witnessed participation from technology innovators from fields of AI, data science, analytics, digital art, and more. It included more than 2,400 contributors from 74 countries, garnering 30,000+ views during the livestream sessions.
The hybrid conference was also held in-person in Vancouver after two years of meeting virtually. The conference featured speakers from big tech companies and institutions including MIT’s Sougwen Chung, Pixar’s co-founder Ed Catmull, Stanford University’s Pat Hanrahan, Xbox’s Corporate Vice President Sarah Bond, and Facebook’s Head of Product Ime Archibong.
The event showcased about 240+ technical papers related to various advancements and scientific contributions, alongside roundtable discussions, workshops, paper discussions and others. Besides these, metaverse was also discussed at length. The participants got to experience AR, VR and mixed reality innovations. Metaverse workshops presented current and future possibilities in the virtual universe in conversations about interoperability, 3D modelling and other ideas.
Metaverse workshops also included topics of how computer graphics deals with human bodily functions like the sense of touch with physical hardware in a spatial environment with programs including Apply Hour, Art Papers, Birds of a Feather, Immersive Pavilion inclusive of Web3D ecosystem and solutions for AR and VR.
A session with AWS Innovators included talks and panel discussions from Animal Logic, Bardel Animation, Framestore, FuzzyPixel, Preymaker, Nvidia, Scanline, Pixomondo, Unity and conversations with ‘Love, Death + Robots’ creator and Blur Studios founder Tim Miller.
New technologies unveiled
A recently launched startup out of London, Move.ai partnered with Nvidia to develop an AI-powered system to shift from traditional motion capture process to high-fidelity mocap that can be created from regular video footage. The new system allows pose estimation tools and difficult-to-observe joints detection which allows the seamless capture of motion displayed in the form of a soccer game footage with several players on the field, all being fully motion captured. For indie developers, a lower-end iPhone mocap software was also announced.
Arcturus, a company founded by former employees of Google and Pixar, has collaborated with Dell and HP to create a line of individual products with systems for volumetric video capture and post-processing. The system allows game developers and filmmakers to capture the entire body from 360-degrees and insert it into a 3D environment of games and interactive entertainment.
Researchers at University College London created LookOut which is a camera control system designed using consumer-grade video gear mounted with software that allows voice commands to direct gimbal movements. The software also uses advanced computer vision similar to Tesla’s self-driving cars to identify people or objects in front of the camera. Users can define a personalised range of actions for the gimbal to identify and track and also reframe actors and objects in a shot.
Expanding on the earlier creation of Hair-GAN—Generative Adversarial Network, a Deep Learning concept—researchers at Kyushu University in Japan have developed a technique to generate CG recreations of hairstyles using a single image. This will enable developers to groom Photoreal CG hair with greater ease.
What’s new?
Numerous new hardwares were also announced by various big tech companies. For instance, Intel announced new lines of ray-tracing GPUs—Arc Pro A-series, for creator applications like Blender, DaVinci Resolve Studio, Handbrake with AV1 hardware encoding that is suitable for mobile form factors and small desktops and can work for engineering, construction and architectural manufacturing.
Nvidia announced a series of technology innovations like the Nvidia Omniverse Avatar Cloud Engine that offers a platform with tools to create AI-powered virtual assistants. The company also stepped into metaverse applications with their new NeuralVBD library—a next generation OpenVBD open-source library. In addition, Nvidia announced their focus on enhancing open-source Universal Scene Description (USD) format to enable smoother function of metaverse applications.
Nvidia also announced two new SDKs with Kaolin WISP and NeuralVDB that will enable developers to power neural graphics in creation and animation of 3D objects to enable easier presentation of models and significant reduction in memory footprint.
Other highlights
SIGGRAPH honoured this year’s contributor with ‘Best of’ from various programmes like—
Best Paper Awards
- Instant Neural Graphics Primitives With a Multiresolution Hash Encoding
- DeepPhase: Periodic Autoencoders for Learning Motion Phase Manifolds
- Image Features Influence Reaction Time: A Learned Probabilistic Perceptual Model for Saccade Latency
- CLIPasso: Semantically Aware Object Sketching
VR Theatre
- Best in Show: On the Morning You Wake (to the End of the World)
Art Gallery
- Best in Show: Ray
Art Papers
Electronic Theatre
- Best in Show: The Seine’s of Tears
- Jury’s Choice: The End of War
- Best Student Project: Yallah!
- Audience Choice: Alternate Mesozoic
Emerging Technologies
- Best in Show: HDR VR
Immersive Pavilion
- Top Selection: Journal of My Journey: Seamless Interaction in Virtuality and Reality With Digital Fabrication and Sensory Feedback
Real-time Live!
- Best in Show: A Showcase of Decima Engine in Horizon Forbidden West
- Audience Choice: AI and Physics Assisted Character Pose Authoring
- OpenAI Says AGI is Here. But Not Everyone is Buying It
- Why Oracle is Bringing Post-Quantum Security to Java
- What Really Happens When You Turn Your Selfie Into a 1980s AI Pic?
- AI Agents Are Making Data Security A Bigger Enterprise Problem
- OpenAI is Getting Nervous About Reinforcement Learning
- Meta’s Bold Open-Weight Bet is Meant to Muse Developers