Shutterstock announced plans to sign collaboration deals with Luma Labs AI, RECON Labs, and Volinga AI, three leading startups in emerging neural radiance fields (NeRF) technology. The companies reported their respective plans at SIGGRAPH 2023, the 50th Annual International Conference and Exhibition on Computer Graphics and Interactive Techniques.
“We are incredibly excited to expand our research and development efforts into this next phase of 3D AI technology and to do so in collaboration with some of the most innovative and forward-thinking minds in creative production,” said Dade Orgeron, Vice President of Innovation at Shutterstock/Turbosquid. “What’s critically important to us as we work to evolve awareness and adoption of NeRF technology is working with innovators who share our commitment to advancing creative technology that hyper-enables the creative process while also protecting and compensating our artist community.”
The Corrider Crew produced an excellent NeRF primer if you are new to the term or just need a refresher:
With both Luma and RECON Labs’ 3Dpresso, Shutterstock intends to explore the rapidly developing field of NeRF technology using Shutterstock’s extensive 2D/3D asset database and adding each companies’ existing 3D assets to Shutterstock’s platform, Turbosquid, for commercial licensing.
“3D is a powerful creative medium that will be a leading format for stories of the future, which is why we founded Luma in 2021,” said Amit Jain, Cofounder and CEO of Luma. “We are on a mission to democratize 3D and make Hollywood quality, photorealistic 3D accessible to everyone. Realtime, high quality, and lifelike 3D scenes are now a reality for everyone using Luma, whether on the web or in Unreal Engine with the new .luma scenes. As the world’s leading NeRF research and product company, we are excited to explore a partnership with Shutterstock and bring Luma’s best-in-class 3D generative platform to Shutterstock’s creators. This will further accelerate the arrival of this incredible 3D technology to every artist, every studio, and every phone!”
Traditionally, 3D scene creation has taken artists several hours to produce depending on the complexity and resolution for the desired landscape or environment. With the introduction of NeRF technology, built using advanced machine learning and computer vision, 3D creators can produce high-resolution, photorealistic 3D scenes from 2D video in minutes.
“NeRF technology will unlock the opportunity for Shutterstock’s community of video contributors to help build the world’s largest library of NeRF based content, from full environments to individual objects from around the world,” said Orgeron. “NeRF will revolutionize the way we create 3D content, by utilizing the tool photographers and videos are already comfortable with—their cameras.”
Shutterstock has also entered into an agreement with Volinga AI to explore licensing and distribution of NVOL files. Volinga, through its professional NeRF Suite, offers a comprehensive range of NeRF-related products and services to support virtual production, VFX, TV and broadcasting, XR and videogames. As part of the agreement, Shutterstock and Volinga will also explore the creation of an NVOL library of high-quality content for Volinga users and the first network of creators to produce on-demand NVOL environments.
This Volinga network would add to Shutterstock’s existing network of more than two million artists and creators around the world who contribute to the Shutterstock library of ethically-sourced, licensable 2D, 3D, video, and audio content. To date, Shutterstock has compensated hundreds of thousands of artists, with anticipated payments to millions more, through its Contributor Fund for the role their content IP has played in training Shutterstock’s generative technology and has provided artists with ongoing royalties tied to licensing activity for newly generated assets.
“We are on a mission to unleash the full potential of NeRF and make this groundbreaking technology accessible to professionals,” said Fernando Rivas-Manzaneque, Co-Founder of Volinga AI. “This library of production-ready NVOL assets will allow our users to start experimenting and using NeRF in their volumes straight away, lowering the entry barrier for this technology. Furthermore, the network of creators to produce on-demand NVOL environments we are envisioning with Shutterstock will definitely take virtual production to the next level.”
Under the new TurboSquid Labs, where Shutterstock will release the first example of its Virtual Locations product—complete, production ready NeRF environments—the company will also working closely with Luma, Volinga, and RECON Labs, and other innovators in NeRF and generative AI fields, to develop a specification and best practices standard for capturing high quality NeRFs.
Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.