Presentation

<
Img2CADSeq: Image-to-CAD Generation via Sequence-Based Diffusion
SessionLearning for CAD
DescriptionImg2CADSeq is a multi-stage pipeline converting single-view images into parametrically editable 3D CAD models. Using a coarse-to-fine hierarchical codebook and VQ-Diffusion , it bridges 2D visual features with 3D semantics , generating standard STEP files ready for direct commercial software use.
Authors
Shiyu Tan
School of Software and BNRist, Tsinghua University
Zixuan Zhao
School of Software and BNRist, Tsinghua University
Hao Gao
School of Software and BNRist, Tsinghua University
Zhiheng Chen
School of Software and BNRist, Tsinghua University
Xiaolong Yin
School of Software and BNRist, Tsinghua University
Enya Shen
School of Software and BNRist, Tsinghua University
Haihe Lab of ITAI
Event Type
Technical Paper
TimeMonday, 20 July 202611:35am - 11:45am PDT
LocationRoom 403 B
Digital Library PDF
Session Time & Location
Sunday, 19 July 20266:00pm - 8:45pm PDTHall K
Monday, 20 July 202610:45am - 12:15pm PDTRoom 403 B
Keywords
Geometry
Modeling
Registration Categories
Full Conference Supporter
Full Conference
Next PresentationNext PresentationLearning for CAD - Interactive Discussion
Similar Presentations
Flatten the Complex: Joint B-Rep Generation via Compositional k-Cell Particles
BrepForge: Factorized B-rep Synthesis via Wireframe Composition and Boundary-Conditioned Surface Instantiation
PatternGSL: A Structured Specification Language for Template-Free and Simulation-Ready 3D Garments