Editor’s note: This post is part of Into the Omniverse, a blog series focused on how developers, 3D artists and enterprises can transform their workflows using the latest advances in OpenUSD and NVIDIA Omniverse.
3D product configurators are changing the way industries like retail and automotive engage with customers by offering interactive, customizable 3D visualizations of products.
Using physically accurate product digital twins, even non-3D artists can streamline content creation and generate stunning marketing visuals.
With the new NVIDIA Omniverse Blueprint for 3D conditioning for precise visual generative AI, developers can start using the NVIDIA Omniverse platform and Universal Scene Description (OpenUSD) to easily build personalized, on-brand and product-accurate marketing content at scale.
By integrating generative AI into product configurators, developers can optimize operations and reduce production costs. With repetitive tasks automated, teams can focus on the creative aspects of their jobs.
Developing Controllable Generative AI for Content Production
The new Omniverse Blueprint introduces a robust framework for integrating generative AI into 3D workflows to enable precise and controlled asset creation.
Example images created using the NVIDIA Omniverse Blueprint for 3D conditioning for precise visual generative AI.
Key highlights of the blueprint include:
- Model conditioning to ensure that the AI-generated visuals adhere to specific brand requirements like colors and logos.
- Multimodal approach that combines 3D and 2D techniques to offer developers complete control over final visual outputs while ensuring the product’s digital twin remains accurate.
- Key components such as an on-brand hero asset, a simple and untextured 3D scene, and a customizable application built with the Omniverse Kit App Template.
- OpenUSD integration to enhance development of 3D visuals with precise visual generative AI.
- Integration of NVIDIA NIM, such as the Edify 360 NIM, Edify 3D NIM, USD Code NIM and USD Search NIM microservices, allows the blueprint to be extensible and customizable. The microservices are available to preview on build.nvidia.com.
How Developers Are Building AI-Enabled Content Pipelines
Katana Studio developed a content creation tool with OpenUSD called COATcreate that empowers marketing teams to rapidly produce 3D content for automotive advertising. By using 3D data prepared by creative experts and vetted by product specialists in OpenUSD, even users with limited artistic experience can quickly create customized, high-fidelity, on-brand content for any region or use case without adding to production costs.
Global marketing leader WPP has built a generative AI content engine for brand advertising with OpenUSD. The Omniverse Blueprint for precise visual generative AI helped facilitate the integration of controllable generative AI in its content creation tools. Leading global brands like The Coca-Cola Company are already beginning to adopt tools from WPP to accelerate iteration on its creative campaigns at scale.
Watch the replay of a recent livestream with WPP for more on its generative AI- and OpenUSD-enabled workflow:
The NVIDIA creative team developed a reference workflow called CineBuilder on Omniverse that allows companies to use text prompts to generate ads personalized to consumers based on region, weather, time of day, lifestyle and aesthetic preferences.
Developers at independent software vendors and production services agencies are building content creation solutions infused with controllable generative AI and built on OpenUSD. Accenture Song, Collective World, Grip, Monks and WPP are among those adopting Omniverse Blueprints to accelerate development.
Read the tech blog on developing product configurators with OpenUSD and get started developing solutions using the DENZA N7 3D configurator and CineBuilder reference workflow.
Get Plugged Into the World of OpenUSD
Various resources are available to help developers get started building AI-enabled product configuration solutions:
For more on optimizing OpenUSD workflows, explore the new self-paced Learn OpenUSD training curriculum that includes free Deep Learning Institute courses for 3D practitioners and developers. For more resources on OpenUSD, attend our instructor-led Learn OpenUSD courses at SIGGRAPH Asia on December 3, explore the Alliance for OpenUSD forum and visit the AOUSD website.
Don’t miss the CES keynote delivered by NVIDIA founder and CEO Jensen Huang live in Las Vegas on Monday, Jan. 6, at 6:30 p.m. PT for more on the future of AI and graphics.
Stay up to date by subscribing to NVIDIA news, joining the community and following NVIDIA Omniverse on Instagram, LinkedIn, Medium and X.
