TransCAD: A Hierarchical Transformer for CAD Sequence Inference from Point Clouds
Authors:
Preprint
Abstract
3D reverse engineering, in which a CAD model is inferred given a 3D scan of a physical object, is a research direction that offers many promising practical applications. This paper proposes TransCAD, an end-to-end transformer-based architecture that predicts the CAD sequence from a point cloud. TransCAD leverages the structure of CAD sequences by using a hierarchical learning strategy. A loop refiner is also introduced to regress sketch primitive parameters. Rigorous experimentation on the DeepCAD and Fusion360 datasets show that TransCAD achieves state-of-the-art results. The result analysis is supported with a proposed metric for CAD sequence, the mean Average Precision of CAD Sequence, that addresses the limitations of existing metrics.Keywords
All Related Versions
This publication has 28 references indexed in Scilit:
- ExtrudeNet: Unsupervised Inverse Sketch-and-Extrude for Shape Parsing2022
- CADOps-Net: Jointly Learning CAD Operation Types and Steps from Boundary-Representations2022
- Learning Gradient Fields for Shape Generation2020
- ParSeNet: A Parametric Surface Fitting Network for 3D Point Clouds2020
- PQ-NET: A Generative Part Seq2Seq Network for 3D Shapes2020
- Learning Implicit Fields for Generative Shape Modeling2019
- ABC: A Big CAD Model Dataset for Geometric Deep Learning2019
- Supervised Fitting of Geometric Primitives to 3D Point Clouds2019
- Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images2018
- EC-Net: An Edge-Aware Point Set Consolidation Network2018