Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.13293 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916009047228416 |
|---|---|
| author | Tan, Shiyu Zhao, Zixuan Gao, Hao Chen, Zhiheng Yin, Xiaolong Shen, Enya |
| author_facet | Tan, Shiyu Zhao, Zixuan Gao, Hao Chen, Zhiheng Yin, Xiaolong Shen, Enya |
| contents | Boundary Representation (BRep) is the standard format for Computer-Aided Design (CAD), yet reconstructing high-quality BReps from single-view images remains challenging due to the complexity of topological constraints and operation sequences. We present Img2CADSeq, a multi-stage pipeline that overcomes these limitations by encoding CAD sequences into a three-level hierarchical codebook. Guided by an importance prioritization, this strategy values profiles over details, compressing long sequences into a stable discrete latent space. To bridge the modality gap, we leverage a coarse-to-fine point cloud intermediate, aligning 2D visual features with 3D CAD sequences via contrastive learning to condition a VQ-Diffusion model. Supported by newly introduced CAD-220K and PrintCAD datasets, our approach ensures robust industrial domain adaptation. Extensive experiments demonstrate that Img2CADSeq significantly outperforms state-of-the-art methods, producing standard STEP files that can be directly used in commercial CAD software. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_13293 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Img2CADSeq: Image-to-CAD Generation via Sequence-Based Diffusion Tan, Shiyu Zhao, Zixuan Gao, Hao Chen, Zhiheng Yin, Xiaolong Shen, Enya Computer Vision and Pattern Recognition Boundary Representation (BRep) is the standard format for Computer-Aided Design (CAD), yet reconstructing high-quality BReps from single-view images remains challenging due to the complexity of topological constraints and operation sequences. We present Img2CADSeq, a multi-stage pipeline that overcomes these limitations by encoding CAD sequences into a three-level hierarchical codebook. Guided by an importance prioritization, this strategy values profiles over details, compressing long sequences into a stable discrete latent space. To bridge the modality gap, we leverage a coarse-to-fine point cloud intermediate, aligning 2D visual features with 3D CAD sequences via contrastive learning to condition a VQ-Diffusion model. Supported by newly introduced CAD-220K and PrintCAD datasets, our approach ensures robust industrial domain adaptation. Extensive experiments demonstrate that Img2CADSeq significantly outperforms state-of-the-art methods, producing standard STEP files that can be directly used in commercial CAD software. |
| title | Img2CADSeq: Image-to-CAD Generation via Sequence-Based Diffusion |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.13293 |