Deep learning for 3D point clouds by Dr Min Wang - UNSW.ai Workshop

UNSW.ai Workshop - Imaging, Sensing and Data Informatics with AI
Title: Deep learning for 3D point clouds by Dr. Min Wang from the special interest group in AI and Cyber Security, UNSW Canberra

Abstract

3D sensors such as LiDARs, depth cameras and 3D scanners are becoming more and more accessible. The 3D data acquired by these sensors can provide rich geometric, shape and scale information, hence providing another channel, complementary to 2D images, to better understand the environment around the machine. 3D data can be represented in different formats, including point clouds, meshes, depth images and volumetric grids. Among these formats, point cloud representations are preferred and commonly used in many applications since they keep the original geometric information in 3D space without any discretization. Deep learning has gained great success in computer vision, and the interests have been extended to 3D point cloud learning recently. However, deep learning on 3D point clouds faces unique challenges in processing the point clouds. This talk will present deep learning techniques for 3D point cloud analysis, focusing on several popular models.

About the Speaker:

Min Wang is a research fellow at the School of Engineering and Information Technology, University of New South Wales (UNSW), Canberra, Australia. She received her PhD in Computer Science from UNSW in 2020 with a dissertation on “Learning Brain Biometrics” which involves physiological signal processing and deep learning for representation extraction and biometric pattern recognition. Her research interests include biometrics and security, bio-cryptography, pattern recognition, deep learning, information systems, and recently, 3D point cloud analysis. She chairs the special interest group in AI and Cyber Security, UNSW Canberra.