DreamCraft3D++: Efficient Hierarchical 3D Generation with Multi-Plane Reconstruction Model
원문 초록
We introduce DreamCraft3D++, an extension of DreamCraft3D that enables efficient high-quality generation of complex 3D assets. DreamCraft3D++ inherits the multi-stage generation process of DreamCraft3D, but replaces the time-consuming geometry sculpting optimization with a feed-forward multi-plane based reconstruction model, speeding up the process by 1000x. For texture refinement, we propose a training-free IP-Adapter module that is conditioned on the enhanced multi-view images to enhance texture and geometry consistency, providing a 4x faster alternative to DreamCraft3D's DreamBooth fine-tuning. Experiments on diverse datasets demonstrate DreamCraft3D++'s ability to generate creative 3D assets with intricate geometry and realistic 360{\deg} textures, outperforming state-of-the-art image-to-3D methods in quality and speed. The full implementation will be open-sourced to enable new possibilities in 3D content creation.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
PhysX: Physical-Grounded 3D Asset Generation
Objaverse++: Curated 3D Object Dataset with Quality Annotations
MEAT: Multiview Diffusion Model for Human Generation on Megapixels with Mesh Attention
Kiss3DGen: Repurposing Image Diffusion Models for 3D Asset Generation
Tactile DreamFusion: Exploiting Tactile Sensing for 3D Generation
3D-Adapter: Geometry-Consistent Multi-View Diffusion for High-Quality 3D Generation
토론
이 논문과 연결된 포럼 글이 아직 없습니다.