Timezone: »

PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
Charles Ruizhongtai Qi · Li Yi · Hao Su · Leonidas Guibas

Mon Dec 04 06:30 PM -- 10:30 PM (PST) @ Pacific Ballroom #13

Few prior works study deep learning on point sets. PointNet is a pioneer in this direction. However, by design PointNet does not capture local structures induced by the metric space points live in, limiting its ability to recognize fine-grained patterns and generalizability to complex scenes. In this work, we introduce a hierarchical neural network that applies PointNet recursively on a nested partitioning of the input point set. By exploiting metric space distances, our network is able to learn local features with increasing contextual scales. With further observation that point sets are usually sampled with varying densities, which results in greatly decreased performance for networks trained on uniform densities, we propose novel set learning layers to adaptively combine features from multiple scales. Experiments show that our network called PointNet++ is able to learn deep point set features efficiently and robustly. In particular, results significantly better than state-of-the-art have been obtained on challenging benchmarks of 3D point clouds.

Author Information

Charles Ruizhongtai Qi (Stanford University)

Charles Ruizhongtai Qi is currently a PhD student at Stanford University. He is a member in Geometric Computing Group and Artificial Intelligence Lab. His research focuses on computer vision and machine learning. Specifically he works on connecting 2D images and 3D shapes as well as 3D deep learning for semantic understanding. Prior to joining Stanford, he got his bachelor degree in Electronic Engineering from Tsinghua University in 2013, with an outstanding graduate award. He has also been an exchange student in Aalto University, Helsinki during 2011 autumn. During 2016 summer, he was a software engineer intern at Google's self-driving car team.

Li Yi (Stanford University)
Hao Su (Stanford)
Leonidas Guibas (stanford.edu)

More from the Same Authors