2 Dimensional Diffusion

Two-dimensional diffusion models, initially designed for image generation, are being extensively repurposed for 3D content creation. Current research focuses on leveraging pre-trained 2D diffusion models to generate consistent multi-view images, which are then used to reconstruct 3D models, often employing neural radiance fields (NeRFs) or other implicit representations. This approach avoids the need for large 3D training datasets, significantly reducing computational costs and accelerating progress in 3D generation. The resulting techniques are finding applications in diverse fields, including architectural design, avatar creation, and even 3D medical image reconstruction.