CADDreamer: CAD object Generation from Single-view Images

Yuan Li1, Cheng Lin2 , Yuan Liu3, Xiaoxiao Long4, Chenxu Zhang6, Ningna Wang1
Xin Li5 Wenping Wang5 Xiaohu Guo1
1The University of Texas at Dallas, 2The University of Hong Kong, 3The Hong Kong University of Science and Technology, 4Nanjing University, 5Texas A&M University 6Bytedance Corresponding Author

CVPR 2025

method

Pipeline of our CADDreamer

Abstract

Diffusion-based 3D generation has made remarkable progress in recent years. However, existing 3D generative models often produce overly dense and unstructured meshes, which stand in stark contrast to the compact, structured, and sharply-edged Computer-Aided Design (CAD) models crafted by human designers. To address this gap, we introduce CADDreamer, a novel approach for generating boundary representations (B-rep) of CAD objects from a single image. CADDreamer employs a primitive-aware multi-view diffusion model that captures both local geometric details and high-level structural semantics during the generation process. By encoding primitive semantics into the color domain, the method leverages the strong priors of pre-trained diffusion models to align with well-defined primitives. This enables the inference of multi-view normal maps and semantic maps from a single image, facilitating the reconstruction of a mesh with primitive labels. Furthermore, we introduce geometric optimization techniques and topology-preserving extraction methods to mitigate noise and distortion in the generated primitives. These enhancements result in a complete and seamless B-rep of the CAD model. Experimental results demonstrate that our method effectively recovers high-quality CAD objects from single-view images. Compared to existing 3D generation techniques, the B-rep models produced by CADDreamer are compact in representation, clear in structure, sharp in edges, and watertight in topology.



Comparison with Other Generation Methods on Synthetic data


method

The segmentation results and reconstructed B-rep models on synthetic image. Red represents segmentation boundaries and edges of B-rep objects. Other colors represents different primitives: plane , cylinder , sphere , torus , and cone .

Comparison with Other Generation Methods on Real data


method

The segmentation results and reconstructed B-rep models on real image. Red represents segmentation boundaries and edges of B-rep objects. Other colors represents different primitives: plane , cylinder , sphere , torus , and cone .

Related Links.

Wonder3D: Single Image to 3D using Cross-Domain Diffusion

RANSAC: Efficient RANSAC for Point Cloud Shape Detection

SyncDreamer: Generating Multiview-consistent Images from a Single-view Image

Similar to Wonder3D, we input multi-view normal maps into the NeuS to achieve high-quality 3D shape reconstruction. Since texture reconstruction is beyond the scope of our research, we have removed the color inputs and associated loss functions.

BibTeX

If you find this work helpful, you can cite our paper as follows:


            @inproceedings{yuan2025CADDreamer,
                author    = {Yuan Li and Cheng Lin and Yuan Liu and Xiaoxiao Long and Chenxu Zhang and Ningna Wang and Xin Li and Wenping Wang and Xiaohu Guo},
                title     = {CADDreamer: CAD Object Generation from Single-view Images},
                booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
                year      = {2025},
                publisher  = {IEEE},
            }
    

If you have any questions or feedbacks, please contact Yuan Li (Li.Yuan@utdallas.edu).