3D Point Cloud Classification and Segmentation using 3D Modified Fisher\n Vector Representation for Convolutional Neural Networks

2017/11/22 by Yizhak Ben-Shabat, Ben-Shabat, Yizhak, Michael Lindenbaum +3 · 2 citations
Earth and Planetary Sciences · Engineering · Environmental Science · #3D Shape Modeling and Analysis #3D Surveying and Cultural Heritage #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Remote Sensing and LiDAR Applications

paper · pdf · doi:10.48550/arxiv.1711.08241

openalex publication_date 2017/11/22 · openalex created_date 2022/08/29 · openalex updated_date 2026/07/28

Abstract

The point cloud is gaining prominence as a method for representing 3D shapes,\nbut its irregular format poses a challenge for deep learning methods. The\ncommon solution of transforming the data into a 3D voxel grid introduces its\nown challenges, mainly large memory size. In this paper we propose a novel 3D\npoint cloud representation called 3D Modified Fisher Vectors (3DmFV). Our\nrepresentation is hybrid as it combines the discrete structure of a grid with\ncontinuous generalization of Fisher vectors, in a compact and computationally\nefficient way. Using the grid enables us to design a new CNN architecture for\npoint cloud classification and part segmentation. In a series of experiments we\ndemonstrate competitive performance or even better than state-of-the-art on\nchallenging benchmark datasets.\n

Citations

• A scalable active framework for region annotation in 3D shape collections
• A Point Set Generation Network for 3D Object Reconstruction from a Single Image
• Multi-view Convolutional Neural Networks for 3D Shape Recognition

Cited by

• A survey of Object Classification and Detection based on 2D/3D data
• Go Wider: An Efficient Neural Network for Point Cloud Analysis via Group Convolutions

Related

• Classification of Point Cloud Scenes with Multiscale Voxel Deep Network
• Deep Learning for 3D Point Cloud Understanding: A Survey
• Clustering based Point Cloud Representation Learning for 3D Analysis
• Deep Learning for 3D Point Clouds: A Survey
• Deep Learning for 3D Point Clouds: A Survey