Pointer-CAD v2: Plan-Then-Construct CAD Generation with Dimension-Aware Parametric Precision

2026-06-28

AI Summary

Existing CAD generation methods prioritize visual similarity over geometric precision and suffer from quantization-induced errors in parametric representations, rendering them inadequate for industrial applications demanding exact dimensional accuracy. To address this, this work proposes a Plan-Then-Construct paradigm: first generating a structured design plan with explicit continuous dimensional parameters, then employing a pointer mechanism during the construction phase to directly reference these parameters, thereby enabling high-fidelity modeling. This approach achieves the first end-to-end prediction of continuous parameters, eliminating quantization errors, and ensures dimensional consistency by decoupling parameter inference from geometric construction. We introduce the first large-scale CAD dataset annotated with design plans and propose a three-level geometric evaluation metric—vertices, edges, and faces. Experiments demonstrate that our method significantly outperforms existing approaches across all geometric levels, making it suitable for precision-sensitive engineering tasks.

Abstract

Computer-aided design (CAD) plays a fundamental role in modern manufacturing by providing the high precision required for industrial production. Recent large language model based approaches formulate CAD generation as a sequence prediction problem and have achieved promising results. However, existing methods and evaluation protocols primarily emphasize visual similarity, while overlooking precise geometric parameters and correct metric scale. Small numerical deviations that are negligible at the shape-level may still violate industrial tolerance requirements, a problem further compounded by current autoregressive paradigms that utilize command sequence representations, aggressively quantize numerical parameters to ease LLM prediction. In this work, we present Pointer-CAD v2. Compared with v1 (arXiv:2603.04337), this version directly predicts continuous values, bypassing the need for quantized numerical parameters and thereby eliminating quantization errors. Specifically, we propose a unified framework that decouples parameter reasoning from geometric construction through a Plan-Then-Construct paradigm. Our method first produces a structured design plan with explicit metric scale parameters. These parameters are organized into a dictionary and directly referenced during sequence generation via a pointer mechanism, eliminating discretization errors and ensuring dimensionally consistent execution. In addition, we construct a new large-scale dataset with plan-level annotation and introduce three hierarchical geometry accuracy metrics to evaluate parametric fidelity at the vertex, edge, and face levels. Extensive experiments demonstrate that Pointer-CAD v2 consistently outperforms existing baselines and achieves substantial improvements in geometric accuracy, enabling reliable CAD generation for precision-critical engineering applications.

Problem

• CAD generation
• geometric accuracy
• parametric precision
• quantization error
• industrial tolerance

Innovation

• Plan-Then-Construct
• dimension-aware
• pointer mechanism
• continuous parameter prediction
• parametric fidelity