RandLANet
About
Category | Paris-Lille-3D |
Short name | RandLANet |
Long name | RandLA-Net |
URL | https://github.com/QingyongHu/RandLA-Net |
Description | |
BibTeX | @article{hu2019randla, title={RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds}, author={Hu, Qingyong and Yang, Bo and Xie, Linhai and Rosa, Stefano and Guo, Yulan and Wang, Zhihua and Trigoni, Niki and Markham, Andrew}, journal={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition}, year={2020} } |
Hardware | AMD Ryzen X3700 GPU 2080Ti |
Runtime | 120 |
Used additional training data | No |
Source code | Yes |
Created at | 08/04/2020 13:48 |
Updated at | 08/04/2020 13:48 |
Result
Av. IoU | Ground | Building | Pole | Bollard | Trash_can | Barrier | Pedestrian | Car | Natural | |
---|---|---|---|---|---|---|---|---|---|---|
Result | 78.5 | 99.5 | 97.0 | 71.0 | 86.7 | 50.5 | 65.5 | 49.1 | 95.3 | 91.7 |
Rank | 10 | 6 | 6 | 9 | 4 | 12 | 5 | 15 | 4 | 7 |
Full confusion matrix
L1' | L2' | L3' | L4' | L5' | L6' | L7' | L8' | L9' | |
---|---|---|---|---|---|---|---|---|---|
L1 | 18087399 | 8006 | 545 | 934 | 547 | 9619 | 65 | 2682 | 8283 |
L2 | 33429 | 6686126 | 630 | 0 | 86 | 11079 | 0 | 0 | 9912 |
L3 | 558 | 6996 | 120634 | 34 | 0 | 252 | 0 | 1354 | 26171 |
L4 | 355 | 0 | 0 | 14432 | 8 | 56 | 0 | 0 | 64 |
L5 | 201 | 0 | 0 | 402 | 20086 | 9716 | 0 | 3211 | 244 |
L6 | 11998 | 19358 | 60 | 367 | 4974 | 244720 | 0 | 2193 | 26476 |
L7 | 314 | 6686 | 357 | 0 | 84 | 790 | 13162 | 1471 | 3690 |
L8 | 2608 | 8763 | 107 | 0 | 0 | 4690 | 187 | 916131 | 17149 |
L9 | 4577 | 100480 | 12226 | 0 | 201 | 27225 | 0 | 1306 | 2635629 |