Evolving Evocative 2D Views of Generated 3D Objects
This addresses the challenge of creating coherent 3D and 2D visual content from text for applications in computer graphics and AI, though it appears incremental as it builds on existing CLIP and ImageNet methods.
The paper tackles the problem of generating 3D objects and corresponding 2D renders from text captions, using guidance from ImageNet and CLIP models, with results showing it can produce anamorphic objects that are visually appealing and evoke the target caption.
We present a method for jointly generating 3D models of objects and 2D renders at different viewing angles, with the process guided by ImageNet and CLIP -based models. Our results indicate that it can generate anamorphic objects, with renders that both evoke the target caption and look visually appealing.