At SIGGRAPH 2024, Shutterstock launched a new multi-modal tool that generates 3D assets from text and image prompts, including ready-to-edit meshes, UVs, and 4K PBR materials, with an immediate focus on servicing the needs of Enterprise-sized businesses, via an API. HP and WPP are early adopters.
“In September, we will release a more open self-serve option, which is going to be a credit-based system, starting at a really low price so people can start to try to use it. It’s not something that’s going to live on the TurboSquid site or live on the Shutterstock site as an e-comm experience.”
Dade Orgeron, VP of Innovation, Shutterstock
Assets, lighting and backgrounds can be generated in under 10 seconds before being imported into Blender or Maya through a full floating point 360 HDRI Edify tool and used for prototyping, set dressing and other pre-production tasks. Output formats are currently OBJ, GLB, PLY and USDZ.
“If you want to feed it an image, you can feed it an image, or if you want to feed it a prompt, you can feed it a prompt. It’s your choice on how you want to start. It’ll accept multiple inputs.”
Dade Orgeron, VP of Innovation, Shutterstock
Shutterstock are keen to make it known that all of the models are ethically sourced. The machine learning model was trained on more than 500 million ethically sourced 3D models and 650 million images. They say that the system also prevents the generation of assets that infringe existing IP.
Source: FXGuiide
Inline image ℅ Radisnce Fields