Year: 2020
Type: dissertation
Abstract: The development of practical applications, such as autonomous driving and robotics, has brought increasing attention to 3D point cloud understanding. However, while deep learning methods obtained remarkable success in 2D image tasks, deep models on point clouds still suffer from unique challenges in processing unstructured points with deep neural networks. This thesis reviews milestones and recent progress in different areas of point cloud learning, and proposes a uniform toolbox to help performance evaluation across models.
Author: Haoming Lu
Institution: University of Illinois Urbana-Champaign
Language: English
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3180621285
