Fondant

This is a unique role at the intersection of computer graphics and generative AI. AI has become really good at generating natural language, images and video. But a foundation model for 3D doesn’t exist, likely because there isn’t enough high-quality data to train a large model. The quality bar of 3D assets for games, animation and VFX is really high. Well-topologized geometry and good surface parameterization (UVs) is table stakes for downstream rigging, animation and texturing. At Fondant, we are building tools that combine SOTA computer graphics with generative AI for creating assets that can be animated and rigged. In this role you would be working with combining diffusion models with geometry processing, deformation, surface parameterization and skinning for high-quality animatable 3D assets. If you are familiar with one or more of the above concepts, reach out at my username [at] fondant.design with a short blurb about yourself and link(s) to personal contribution(s) e.g., open-source code, cool demo of something you built etc.