Title:
METHODS AND SYSTEMS FOR SEMANTIC SEGMENTATION OF POINT CLOUD
Document Type and Number:
WIPO Patent Application WO/2023/284251
Kind Code:
A1
Abstract:
Systems, methods and apparatus for sematic segmentation of 3D point clouds using deep neural networks. The deep neural network generally has two primary subsystems: a multi-branch cascaded subnetwork that includes an encoder and a decoder, and is configured to receive a sparse 3D point cloud, and capture and fuse spatial feature information in the sparse 3D point cloud at multiple scales and multi hierarchical levels; and a spatial feature transformer subnetwork that is configured to transform the cascaded features generated by the multi-branch cascaded subnetwork and fuse these scaled features using a shared decoder attention framework to assist in the prediction of sematic classes for the sparse 3D point cloud.
Inventors:
CHENG RAN (CA)
RAZANI RYAN (CA)
LIU BINGBING (CA)
RAZANI RYAN (CA)
LIU BINGBING (CA)
Application Number:
PCT/CN2021/139358
Publication Date:
January 19, 2023
Filing Date:
December 17, 2021
Export Citation:
Assignee:
HUAWEI TECH CO LTD (CN)
International Classes:
G06T7/11
Foreign References:
| US20210082181A1 | 2021-03-18 | |||
| US20210042557A1 | 2021-02-11 | |||
| US10970518B1 | 2021-04-06 | |||
| CN108665496A | 2018-10-16 | |||
| CN111311611A | 2020-06-19 |
Download PDF:
Previous Patent: AEROSOL GENERATING APPARATUS
Next Patent: TUNNEL FIBER COMPOSITE RESIN PAVEMENT STRUCTURE
Next Patent: TUNNEL FIBER COMPOSITE RESIN PAVEMENT STRUCTURE