Summary of Gem3d: Generative Medial Abstractions For 3d Shape Synthesis, by Dmitry Petrov et al.
GEM3D: GEnerative Medial Abstractions for 3D Shape Synthesis
by Dmitry Petrov, Pradyumn Goyal, Vikas Thamizharasan, Vladimir G. Kim, Matheus Gadelha, Melinos Averkiou, Siddhartha Chaudhuri, Evangelos Kalogerakis
First submitted to arxiv on: 26 Feb 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Graphics (cs.GR); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed GEM3D model is a novel deep learning approach for generating 3D shapes that takes into account both topology and geometry. The method represents shapes using a neural skeleton-based framework, which encodes information about the Medial Axis Transform (MAT) and then generates surfaces through a neural implicit formulation that considers topological and geometric constraints. This allows for more accurate surface reconstruction and diverse shape generation compared to previous methods. The authors demonstrate the effectiveness of GEM3D in tasks such as shape synthesis and point cloud reconstruction, achieving state-of-the-art results on challenging datasets like Thingi10K and ShapeNet. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary GEM3D is a new way to create 3D shapes using computers. It’s based on a special kind of skeleton that helps the computer understand what the shape looks like and how it’s connected. This allows the computer to generate really accurate surfaces and create lots of different shapes. The authors tested GEM3D with some tricky tasks, like recreating complex shapes from real-world data. Their results were better than other methods they tried! |
Keywords
* Artificial intelligence * Deep learning
