Unsupervised point cloud feature learning method and device based on local global bidirectional reasoning

Abstract

The invention discloses an unsupervised point cloud feature learning method and device based on local global bidirectional reasoning, and the method comprises the following steps: obtaining three-dimensional point cloud data, carrying out the feature extraction of the three-dimensional point cloud data through a preset neural extraction network, and obtaining a local point cloud feature and a global point cloud feature corresponding to the three-dimensional point cloud data; processing the global point cloud features through a decoder network to obtain an original point cloud and an estimatedpoint cloud normal vector; and performing unsupervised point cloud feature learning according to the original point cloud, the estimated point cloud normal vector and the local point cloud features toobtain network parameters, and generating a target classifier according to the network parameters. Therefore, structural information and semantic knowledge can be learned at the same time, the quality of unsupervised learning features is improved, and therefore the performance of unsupervised point cloud recognition is improved.

Description

technical field [0001] The invention relates to computer three-dimensional point cloud recognition and feature learning technology, in particular to an unsupervised point cloud feature learning method and device based on local and global two-way reasoning. Background technique [0002] In recent years, deep learning methods for point 3D cloud processing have attracted great attention from researchers. Point cloud recognition has great application value in many real-world scenarios, such as autonomous driving, augmented reality, and intelligent robots. A core problem in point cloud recognition is to learn powerful features that are both distinguishable, versatile and robust. In order to solve this problem, existing high-performance point cloud recognition systems usually need to rely on a large amount of manually labeled data, but Manually annotated data often requires high labor costs and may limit the generalization ability of learned models. Therefore, unsupervised featu...

AI Technical Summary

Problems solved by technology

These methods have been proven to be effective in extracting the basic structural information of point clouds, but usually fail to learn high-level semantic information from point clouds

Therefore, the performance of current unsupervised models still lags far behind the state-of-the-art supervised models

Images