Soslan Kabisov, Gennadiy Savrasov
Computer-Aided Design (CAD) is essential in modern engineering, yet the conversion of existing shapes into editable CAD models remains a significant challenge requiring expert intervention. This paper introduces CADENA, a novel approach that reconstructs a 3D mesh as a parametric CAD program by sequentially reversing the order of operations while constantly assessing the target geometry against the predictions. In addition, we present CADENA-Bench, a benchmark designed for evaluating reverse-engineering methods specifically on mechanical parts, which addresses the current lack of such benchmarks. The results demonstrate that CADENA surpasses previous methods in performance, as evidenced by its outcomes on CADENA-Bench, along with established datasets such as DeepCAD, Fusion 360, and MCB. This advancement not only promotes efficiency in CAD model generation from raw geometries but also enhances the replicability of engineering processes. Code, model weights, and the CADENA-Bench are available for public access, facilitating further research and development in this domain.
@article{858aeb53-8339-4081-a864-bac1944c6f68,
title={CADENA: STEPWISE CAD REVERSE ENGINEERING},
author={Soslan Kabisov and Gennadiy Savrasov},
year={2026},
language={en}
}TY - JOUR TI - CADENA: STEPWISE CAD REVERSE ENGINEERING AU - Soslan Kabisov AU - Gennadiy Savrasov PY - 2026 LA - en ER -
This paper addresses the challenge of assessing the feasibility of wind power plant projects at sites with insufficient or no local historic wind data
Important advances in electrochemical engineering technology over the last three decades have fostered the development of a lternative methods to alle
Increasing volumes of waste printed circuit boards from obsolete electronic equipment posed escalating environmental risks and resource losses due to