公开(公告)号：US11810359B2

公开(公告)日：2023-11-07

申请号：US17557933

申请日：2021-12-21

Applicant: DALIAN UNIVERSITY OF TECHNOLOGY

Inventor： Xin Yang ,  Xiaopeng Wei ,  Yu Qiao ,  Qiang Zhang ,  Baocai Yin ,  Haiyin Piao ,  Zhenjun Du

IPC: G06V10/00 ,  G06V20/40 ,  G06V10/46 ,  G06V10/82 ,  G06T3/40 ,  G06T7/215 ,  G06T9/00 ,  G06V10/72 ,  G06V10/764 ,  G06V10/778 ,  G06V10/774 ,  G06V10/776 ,  G06T7/10 ,  G06F18/21 ,  G06F18/214

CPC classification number: G06V20/49 ,  G06F18/217 ,  G06F18/2155 ,  G06T3/4007 ,  G06T3/4046 ,  G06T7/10 ,  G06T7/215 ,  G06T9/002 ,  G06V10/46 ,  G06V10/72 ,  G06V10/764 ,  G06V10/776 ,  G06V10/778 ,  G06V10/7753 ,  G06V10/82 ,  G06V20/41 ,  G06T2207/10016 ,  G06T2207/10024 ,  G06T2207/20081 ,  G06T2207/20084

Abstract: The present invention belongs to the technical field of computer vision, and provides a video semantic segmentation method based on active learning, comprising an image semantic segmentation module, a data selection module based on the active learning and a label propagation module. The image semantic segmentation module is responsible for segmenting image results and extracting high-level features required by the data selection module; the data selection module selects a data subset with rich information at an image level, and selects pixel blocks to be labeled at a pixel level; and the label propagation module realizes migration from image to video tasks and completes the segmentation result of a video quickly to obtain weakly-supervised data. The present invention can rapidly generate weakly-supervised data sets, reduce the cost of manufacture of the data and optimize the performance of a semantic segmentation network.

公开(公告)号：US11875583B2

公开(公告)日：2024-01-16

申请号：US17533878

申请日：2021-11-23

Applicant: DALIAN UNIVERSITY OF TECHNOLOGY

Inventor： Xin Yang ,  Tong Li ,  Baocai Yin ,  Zhaoxuan Zhang ,  Boyan Wei ,  Zhenjun Du

IPC: G06T7/50 ,  G06V20/64 ,  G06T3/00

CPC classification number: G06V20/647 ,  G06T3/0062 ,  G06T7/50 ,  G06T2207/10028

Abstract: The present invention belongs to the technical field of 3D reconstruction in the field of computer vision, and provides a dataset generation method for self-supervised learning scene point cloud completion based on panoramas. Pairs of incomplete point cloud and target point cloud with RGB information and normal information can be generated by taking RGB panoramas, depth panoramas and normal panoramas in the same view as input for constructing a self-supervised learning dataset for training of the scene point cloud completion network. The key points of the present invention are occlusion prediction and equirectangular projection based on view conversion, and processing of the stripe problem and point-to-point occlusion problem during conversion. The method of the present invention includes simplification of the collection mode of the point cloud data in a real scene; occlusion prediction idea of view conversion; and design of view selection strategy.