Submitted by Yizheng Wu.

Submission data

Full nameSelf-Training for Semi-Supervised Learning
DescriptionExploiting instance consistency regularization to boost semi-supervised 3D instance segmentation
Publication titleInstance Consistency Regularization for Semi-Supervised 3D Instance Segmentation
Publication authorsYizheng Wu, Zhiyu Pan, Kewei Wang, Xingyi Li, Jiahao Cui, Liwen Xiao, Guosheng Lin, Zhiguo Cao
Publication venueIEEE Transactions on Pattern Analysis and Machine Intelligence
Publication URLhttps://ieeexplore.ieee.org/abstract/document/10587133
Input Data TypesUses Color        Uses 3D
Programming language(s)Python, CUDA
HardwareRTX 3090
Websitehttps://github.com/W1zheng/InsTeacher3D
Source code or download URLhttps://github.com/W1zheng/InsTeacher3D
Submission creation date6 Dec, 2022
Last edited21 Jul, 2024

3D semantic instance results with limited reconstructions



Infoavg ap 50%bathtubbedbookshelfcabinetchaircountercurtaindeskdoorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwindow
0.4431.0000.6170.3410.3820.7850.0000.3330.1580.4850.4580.4200.2500.0000.3840.6300.4670.8750.394