Submitted anonymously.

Submission data

Full nameRandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds
DescriptionLearning Semantic Segmentation of Large-Scale Point Clouds with Random Sampling
Input Data TypesUses Color,Uses Geometry        Uses 3D
Programming language(s)Tensorflow
HardwareAMD Ryzen X3700 GPU 2080Ti
Submission creation date6 Dec, 2020
Last edited6 Jan, 2021

3D semantic label results

Infoavg ioubathtubbedbookshelfcabinetchaircountercurtaindeskdoorfloorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwallwindow
permissive0.6450.7780.7310.6990.5770.8290.4460.7360.4770.5230.9450.4540.2690.4840.7490.6180.7380.5990.8270.7920.621