ABC: Model Datasets for Geometric Deep Learning
A collection of 1 Million Computer-Aided Design (CAD) Models for Geometric Deep Learning Research
This research summary is just one of many that are distributed weekly on the AI scholar newsletter. To start receiving the weekly newsletter, sign up here.
Big data and deep learning networks are transforming many areas of machine learning (ML). To achieve more accurate models though, adequate data is crucial. Good news — data availability is increasingly being made possible by the ubiquity of acquisition devices and massive data sharing on social media.
However, the situation is completely different for 3D geometric models. Even with growing 3D sensors availability and improved 3D design tools, acquiring or constructing high-quality geometric techniques is still difficult.
A Big CAD Model Dataset For Geometric Deep Learning
Scholars have introduced a new and massive ABC-Dataset comprising a collection of 1 million Computer-Aided Design (CAD) models for research of geometric deep learning methods and applications.
Each model is represented through a collection of plainly parametrized curves and surfaces that provide ground truth for differential quantities, patch segmentation, geometric feature detection, and shape reconstruction.
The researchers have performed a large-scale benchmark for the estimation of surface normals, comparing existing data-driven methods and evaluating their performance against both the ground truth and traditional normal estimation methods.
Potential Uses and Effects
With so much data on the table, its time for the AI community to dig deep into geometric deep learning research and applications. The available 3D dataset can be implemented for data-driven processing and applications of geometrical data.
The dataset is available here
Read more: https://arxiv.org/abs/1812.06216v2
Thanks for reading. Please comment, share and remember to subscribe to our weekly newsletter for the most recent and interesting research papers! You can also follow me on Twitter and LinkedIn. Remember to 👏 if you enjoyed this article. Cheers!