Nvidia and Shutterstock Team Up to Bring Generative AI to 3D Modeling

Nvidia is taking generative AI into a new dimension, literally, by partnering with Shutterstock to create a tool that generates 3D models using AI. This collaboration, announced at Siggraph 2024, brings the power of generative AI to the world of 3D asset creation. While there are other AI tools capable of generating 3D models, Shutterstock’s offering stands out as the most official and integrated solution. The generated models will reside within Shutterstock’s TurboSquid library, joining the existing collection of 3D assets. To create these models, users can provide text and image prompts, allowing for greater creative control and customization. Designers can then edit and refine these assets in separate applications, with export options available in various popular file formats. Nvidia boasts that designers can see a preview of their generated asset in just 10 seconds, quickly iterating and refining their designs. They can then choose to convert the preview into a high-quality 3D asset, applying physically-based rendering (PBR) materials for a realistic look and feel. Beyond individual assets, Nvidia states that the tool can generate 360-degree HDRi backgrounds, capable of simulating realistic lighting within virtual scenes. Nvidia emphasizes that the training data for this generative AI tool comes from licensed, ethically sourced material, ensuring the generated assets are safe for commercial use. While the origins of the dataset are generally reliable, concerns about potential biases or unintended consequences remain, as is common with generative AI tools. This partnership builds on Nvidia’s previous collaboration with Getty Images, where they’ve further enhanced the generative AI features. The image generation capabilities have reportedly been improved for speed and quality, now generating images twice as fast and with higher fidelity. Shutterstock’s 3D feature is currently in beta testing and available exclusively to enterprise customers. Although the company has not confirmed plans for wider accessibility, the potential to extend the tool to non-commercial users exists. At Siggraph, Shutterstock demonstrated the tool’s integration with Blender, allowing direct generation of 3D assets within the application. HP also showcased physical prototypes of 3D models created using Shutterstock’s generative AI, highlighting the potential for real-world applications. This collaboration marks a significant step forward in the realm of generative AI, empowering designers and artists with tools to create complex 3D models with greater efficiency and creativity.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top