Enregistré dans:
Détails bibliographiques
Auteurs principaux: Wang, Yinghui, Zhang, Xinyu, Du, Peng
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2510.17157
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912659721420800
author Wang, Yinghui
Zhang, Xinyu
Du, Peng
author_facet Wang, Yinghui
Zhang, Xinyu
Du, Peng
contents Generating editable, parametric CAD models from a single image holds great potential to lower the barriers of industrial concept design. However, current multi-modal large language models (MLLMs) still struggle with accurately inferring 3D geometry from 2D images due to limited spatial reasoning capabilities. We address this limitation by introducing GACO-CAD, a novel two-stage post-training framework. It is designed to achieve a joint objective: simultaneously improving the geometric accuracy of the generated CAD models and encouraging the use of more concise modeling procedures. First, during supervised fine-tuning, we leverage depth and surface normal maps as dense geometric priors, combining them with the RGB image to form a multi-channel input. In the context of single-view reconstruction, these priors provide complementary spatial cues that help the MLLM more reliably recover 3D geometry from 2D observations. Second, during reinforcement learning, we introduce a group length reward that, while preserving high geometric fidelity, promotes the generation of more compact and less redundant parametric modeling sequences. A simple dynamic weighting strategy is adopted to stabilize training. Experiments on the DeepCAD and Fusion360 datasets show that GACO-CAD achieves state-of-the-art performance under the same MLLM backbone, consistently outperforming existing methods in terms of code validity, geometric accuracy, and modeling conciseness.
format Preprint
id arxiv_https___arxiv_org_abs_2510_17157
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GACO-CAD: Geometry-Augmented and Conciseness-Optimized CAD Model Generation from Single Image
Wang, Yinghui
Zhang, Xinyu
Du, Peng
Computer Vision and Pattern Recognition
Artificial Intelligence
Generating editable, parametric CAD models from a single image holds great potential to lower the barriers of industrial concept design. However, current multi-modal large language models (MLLMs) still struggle with accurately inferring 3D geometry from 2D images due to limited spatial reasoning capabilities. We address this limitation by introducing GACO-CAD, a novel two-stage post-training framework. It is designed to achieve a joint objective: simultaneously improving the geometric accuracy of the generated CAD models and encouraging the use of more concise modeling procedures. First, during supervised fine-tuning, we leverage depth and surface normal maps as dense geometric priors, combining them with the RGB image to form a multi-channel input. In the context of single-view reconstruction, these priors provide complementary spatial cues that help the MLLM more reliably recover 3D geometry from 2D observations. Second, during reinforcement learning, we introduce a group length reward that, while preserving high geometric fidelity, promotes the generation of more compact and less redundant parametric modeling sequences. A simple dynamic weighting strategy is adopted to stabilize training. Experiments on the DeepCAD and Fusion360 datasets show that GACO-CAD achieves state-of-the-art performance under the same MLLM backbone, consistently outperforming existing methods in terms of code validity, geometric accuracy, and modeling conciseness.
title GACO-CAD: Geometry-Augmented and Conciseness-Optimized CAD Model Generation from Single Image
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2510.17157