Submitted by Ryan Chu.

Submission data

Full nameTwo-Way Inter-label Self-Training for Semi-supervised 3D Instance Segmentation
Publication titleTWIST: Two-Way Inter-label Self-Training for Semi-supervised 3D Instance Segmentation
Publication authorsRuihang Chu
Publication venueCVPR 2022
Publication URLhttps://openaccess.thecvf.com/content/CVPR2022/papers/Chu_TWIST_Two-Way_Inter-Label_Self-Training_for_Semi-Supervised_3D_Instance_Segmentation_CVPR_2022_paper.pdf
Input Data TypesUses Color        Uses 3D
Programming language(s)python, CUDA
Hardware2080Ti
Submission creation date25 Oct, 2021
Last edited19 Dec, 2023

3D semantic instance results with limited reconstructions



Infoavg ap 50%bathtubbedbookshelfcabinetchaircountercurtaindeskdoorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwindow
0.4810.6670.7600.4680.3130.8020.0080.5290.0980.3640.4110.3480.5000.5710.5040.6460.5300.9440.201