Submitted by Pedro Hermosilla Casajus.

Submission data

Full nameMonte Carlo Convolutional Neural Networks
Publication titleMonte Carlo Convolution for Learning on Non-Uniformly Sampled Point Clouds
Publication authorsP. Hermosilla, T. Ritschel, P.P. Vazquez, A. Vinacua, T. Ropinski
Publication venueSIGGRAPH Asia 2018
Publication URLhttps://arxiv.org/abs/1806.01759
Input Data TypesUses Color        Uses 3D
Programming language(s)Python, C++ and CUDA
HardwareCore i7, Nvidia RTX 2080 Ti, 16 Gb Ram
Websitehttps://github.com/viscom-ulm/MCCNN
Source code or download URLhttps://github.com/viscom-ulm/MCCNN
Submission creation date11 Jun, 2019
Last edited11 Jun, 2019

3D semantic label results

Infoavg ioubathtubbedbookshelfcabinetchaircountercurtaindeskdoorfloorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwallwindow
permissive0.6330.8660.7310.7710.5760.8090.4100.6840.4970.4910.9490.4660.1050.5810.6460.6200.6800.5420.8170.7950.618