Point cloud to CAD models with improved geometric feature accuracy
PrimitiveCAD: An LLM-Based Point-to-CAD Reconstruction with Primitive-Aware Tokenization and Operation Alignment
Graphics
Summary
Turning 3D point clouds into precise CAD models is challenging, especially when capturing detailed shapes. The authors propose a new method called PrimitiveCAD that better recognizes basic shape building blocks and closely matches CAD operations. They train a large language model with special tokens and rewards to improve accuracy and feature preservation. Their approach worked well on benchmark datasets, producing more accurate and detailed CAD reconstructions than before.
What this means in practice
- •For industrial designers: Create more accurate CAD models from scanned or raw 3D point data, speeding up the early design stages.
- •For mechanical engineers: Improve CAD model fidelity in reverse engineering workflows when converting physical objects into editable CAD files.
Authors
Jian Gao, Kailin Bi, Jiamin Xu, Jinlan Xu, Gang Xu
Abstract
Large-model-based point-to-CAD generation holds immense potential for advancing industrial design and enhancing 3D modeling efficiency. However, most existing methods approach the problem as a general point-cloud encoding and token prediction task, neglecting the tokenization and supervision specifically for CAD-related primitives. As a result, these methods often struggle to accurately reconstruct the intricate primitive structures. To address this limitation, we propose PrimitiveCAD, a novel multi-stage paradigm for point-to-CAD reconstruction that enhances the geometric accuracy of generated CAD models while better preserving critical geometric features. First, we introduce a primitive-aware point cloud tokenization model, enabling the system to learn more robust geometric representations from CAD point clouds. Next, we perform supervised finetuning on a large language model (LLM) and introduce an operation alignment loss to align key CAD operation frequencies, thereby improving the preservation of global shape features. Finally, we incorporate reinforcement learning (RL) and introduce a feature-line alignment reward to further reduce stochasticity and enhance the fine-grained preservation of geometric features. Experiments on the DeepCAD and Fusion360 datasets show that our method achieves state-of-the-art performance in code validity, geometric accuracy, and geometric feature preservation.