Problem
Research questions and friction points this paper is trying to address.
Engineering Drawings
SVG Generation
Multi-View Alignment
Geometric Fidelity
Computer-Aided Design
DrawingsDreamer: A Unified Multi-View Engineering Drawings Generation Model
🤖 AI Summary
This work addresses the challenge of preserving geometric fidelity and cross-view spatial alignment in engineering drawings with existing models by proposing a large language model-driven framework for multi-view vector drawing generation. The framework unifies the generation task as sequence modeling, eliminating the need for raster encoders. It introduces a novel hierarchical suffix tokenization representation alongside a progressive curriculum scheduling strategy, enabling a smooth transition from local structural inpainting to macro-level generation. Experimental results demonstrate that the proposed method significantly improves both the geometric fidelity and syntactic accuracy of generated drawings across various conditional and unconditional generation tasks.
📝 Abstract
Scalable Vector Graphics (SVG) are essential for modern industrial Computer-Aided Design (CAD). However, existing autoregressive SVG generation models are predominantly tailored for artistic creation and struggle to maintain the rigorous geometric fidelity and cross-view spatial alignment required for engineering drawings. To bridge this gap, we introduce \textbf{DrawingsDreamer}, a unified Large Language Model (LLM)-driven framework for multi-view vector-based engineering drawings generation. By formulating the generation of multi-view engineering drawings purely as a sequence modeling task, we eliminate the need of raster image encoders. We propose a Streamlined Representation utilizing hierarchical postfix tokenization, which guides the model to establish local geometric coordinates before assigning semantic boundaries. Optimized via a progressive task-aware curriculum schedule, \textbf{DrawingsDreamer} effectively transitions from localized structural repair to macroscopic generation in a unified model. Extensive experiments demonstrate that our unified model achieves strong performance in both geometric fidelity and syntactic accuracy across diverse conditional and unconditional generation tasks.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Multi-View Engineering Drawings
Scalable Vector Graphics (SVG)
Large Language Model (LLM)
Sequence Modeling
Curriculum Learning
🔎 Similar Papers
No similar papers found.
