This repository complements "What's the Situation with Intelligent Mesh Generation: A Survey and Perspectives"[ arxiv ]. To facilitate better access to the relevant literature, we provide links to the 110 articles mentioned in the article by technical categories, as well as the relevant codes for these articles. In addition, we also provide links to commonly used datasets.
- IMG_Survey
Intelligent mesh generation (IMG) is a relatively new field that refers to a kind of method to generate mesh by machine learning. IMG has greatly expanded the generalizability and practicality of mesh generation techniques, bringing many breakthroughs and potential possibilities for mesh generation. Within its short life span, we have seen tremendous advances in this field. However, there is a lack of surveys focusing on IMG methods covering recent works. In this article, we present an overview of the existing IMG methods systematically. Focusing on 110 preliminary studies describing different IMG methods, we conducted a comprehensive analysis and evaluation from the core technique and application scope of the algorithm, agent learning goals, data types, targeting challenges, advantages and limitations. With the aim of literature collection and classification based on content extraction, we provide three different taxonomies from three views of key technique, output mesh unit element, and applicable input data types. Finally, we highlight some promising future research directions and challenges in IMG.
| Paper | Code | Source | connection map |
|---|---|---|---|
| A density driven mesh generator guided by a neural network | -- | IEEE T MAGN 1993 | -- |
| Automatic mesh generation by the let-it-grow neural network | -- | IEEE T MAGN 1996 | -- |
| A finite element mesh generator based on an adaptive neural network | -- | IEEE T MAGN 1998 | -- |
| A neural network generator for tetrahedral meshes | -- | IEEE T MAGN 2003 | -- |
| An Optimized Generator of Finite Element Meshes Based on a Neural Network | -- | IEEE Transactions on Magnetics 2008 | -- |
| Vis2Mesh: Efficient Mesh Reconstruction From Unstructured Point Clouds of Large Scenes With Learned Virtual View Visibility | code | ICCV 2021 | image |
| Differentiable surface triangulation | code | TOG 2021 | -- |
| Learning Delaunay Surface Elements for Mesh Reconstruction | code | CVPR 2021 | image |
| Scalable Surface Reconstruction with Delaunay-Graph Neural Networks | code | Computer Graph Forum 2021 | image |
| Paper | Code | Source | connection map |
|---|---|---|---|
| An ANN-based element extraction method for automatic mesh generation | -- | Expert Systems with Applications 2005 | -- |
| A new unstructured hybrid mesh generation method based on BP-ANN | -- | IPCS 2022 | -- |
| Reinforcement learning for automatic quadrilateral mesh generation: a soft actor-critic approach | -- | arXiv:2203.11203 | -- |
Commonly used: Princeton ModelNet; ShapeNet; TOSCA; COSEG; surface reconstruction benchmark of Berge and Williams; Thingi10K; D-FAUST; Famous; CAD dataset ABC;
Dataset for single image 3D shape modeling: Pix3d;
Facial expression dataset: COMA; MeIn3D;
Human body shapes datasets: MGN; MultiHuman;
Clothed body meshes with real texture: RenderPeople; Axyz; Digit Wardrobe;
Indoor scenes datasets: ScanNet; Scenenet; Matterport3d; Synthetic Rooms;
Commonly used :QuadWild;
Clothed body meshes with real texture: RenderPeople;Axyz;
@ARTICLE{10141677,
author={Lei, Na and Li, Zezeng and Xu, Zebin and Li, Ying and Gu, Xianfeng},
journal={IEEE Transactions on Visualization and Computer Graphics},
title={What's the Situation With Intelligent Mesh Generation: A Survey and Perspectives},
year={2023},
pages={1-20},
doi={10.1109/TVCG.2023.3281781}}