# KDH-CAD: Knowledge-data hybrid CAD learning under data scarcity

arXiv: https://arxiv.org/abs/2606.01702 · HTML: https://mlsanity.com/paper/2606.01702/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 57 references that are in the map.
- First posted on arXiv 1 Jun 2026.

## Builds on (most cited first, 24 of 24)

- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning](https://mlsanity.com/paper/2307.08691/) (2023)
- [ABC: A Big CAD Model Dataset For Geometric Deep Learning](https://mlsanity.com/paper/1812.06216/) (2018)
- [DeepCAD: A Deep Generative Network for Computer-Aided Design Models](https://mlsanity.com/paper/2105.09492/) (2021)
- [VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks](https://mlsanity.com/paper/2410.05160/) (2024)
- [BRepNet: A topological message passing system for solid models](https://mlsanity.com/paper/2104.00706/) (2021)
- [Computer-Aided Design as Language](https://mlsanity.com/paper/2105.02769/) (2021)
- [SolidGen: An Autoregressive Model for Direct B-rep Synthesis](https://mlsanity.com/paper/2203.13944/) (2022)
- [JoinABLe: Learning Bottom-up Assembly of Parametric CAD Joints](https://mlsanity.com/paper/2111.12772/) (2021)
- [SketchGraphs: A Large-Scale Dataset for Modeling Relational Geometry in Computer-Aided Design](https://mlsanity.com/paper/2007.08506/) (2020)
- [ROCA: Robust CAD Model Retrieval and Alignment from a Single Image](https://mlsanity.com/paper/2112.01988/) (2021)
- [Generating CAD Code with Vision-Language Models for 3D Designs](https://mlsanity.com/paper/2410.05340/) (2024)
- [Text-to-CAD Generation Through Infusing Visual Feedback in Large Language Models](https://mlsanity.com/paper/2501.19054/) (2025)
- [Rethinking Confidence Calibration for Failure Prediction](https://mlsanity.com/paper/2303.02970/) (2023)
- [FlexCAD: Unified and Versatile Controllable CAD Generation with Fine-tuned Large Language Models](https://mlsanity.com/paper/2411.05823/) (2024)
- [MeshMAE: Masked Autoencoders for 3D Mesh Data Analysis](https://mlsanity.com/paper/2207.10228/) (2022)
- [3D-PreMise: Can Large Language Models Generate 3D Shapes with Sharp Features and Parametric Control?](https://mlsanity.com/paper/2401.06437/) (2024)
- [CAD-Editor: A Locate-then-Infill Framework with Automated Training Data Synthesis for Text-Based CAD Editing](https://mlsanity.com/paper/2502.03997/) (2025)
- [Bringing Attention to CAD: Boundary Representation Learning via Transformer](https://mlsanity.com/paper/2504.07134/) (2025)
- [Don't Mesh with Me: Generating Constructive Solid Geometry Instead of Meshes by Fine-Tuning a Code-Generation LLM](https://mlsanity.com/paper/2411.15279/) (2024)
- [Intelligent CAD 2.0](https://mlsanity.com/paper/2410.03759/) (2024)
- [Push-pull direct modeling of solid CAD models](https://mlsanity.com/paper/1811.04036/) (2018)
- [Constraint-Aware Feature Learning for Parametric Point Cloud](https://mlsanity.com/paper/2411.07747/) (2024)

## Abstract (arXiv)

Deep learning in computer-aided design (CAD) remains fundamentally constrained by the data scarcity challenge: authentic CAD data is difficult to collect at scale, while synthetic data may not faithfully reflect real design practice. Rather than pursuing ever-larger CAD datasets, this paper alternatively treats CAD learning as a knowledge completion and calibration problem. It introduces KDH-CAD, a knowledge-data hybrid framework that integrates pretrained knowledge in foundation models, structured domain knowledge from textbooks/tutorials, and a very small amount of labeled CAD data. Domain knowledge is used to elicit and complete CAD-relevant concepts that are weakly expressed or under-represented in pretrained foundation models, while labeled CAD data calibrates these concepts in the latent space to account for task-specific geometric variability, without fine-tuning the foundation model. Experiments on real-world mechanical part classification show that KDH-CAD achieves strong performance in low-data regimes, reaching 92.6% accuracy with only 250 training samples, 95.8% with 1,000 samples, and continuing to improve with additional data. This matches or exceeds state-of-the-art performance that typically requires an order of magnitude more data. These results suggest that combining pretrained foundation models with structured domain knowledge can substantially reduce reliance on large-scale CAD datasets, providing a principled and practical direction for data-efficient CAD learning.