Phidias: A Generative Model for Creating 3D Content from Text, Image, and 3D Conditions with Reference-Augmented Diffusion
This addresses the need for more efficient and controlled 3D modeling tools for designers, though it appears incremental as it builds on existing diffusion methods with novel enhancements.
The paper tackles the problem of generating 3D content by using a reference 3D model to guide diffusion-based generation from text, image, and 3D conditions, resulting in improved quality, generalization, and controllability over existing methods.
In 3D modeling, designers often use an existing 3D model as a reference to create new ones. This practice has inspired the development of Phidias, a novel generative model that uses diffusion for reference-augmented 3D generation. Given an image, our method leverages a retrieved or user-provided 3D reference model to guide the generation process, thereby enhancing the generation quality, generalization ability, and controllability. Our model integrates three key components: 1) meta-ControlNet that dynamically modulates the conditioning strength, 2) dynamic reference routing that mitigates misalignment between the input image and 3D reference, and 3) self-reference augmentations that enable self-supervised training with a progressive curriculum. Collectively, these designs result in a clear improvement over existing methods. Phidias establishes a unified framework for 3D generation using text, image, and 3D conditions with versatile applications.