| Short Name | KP-FCNN |
| Long Name | KPConv Segmentation Network for Point Clouds |
| Website | https://github.com/HuguesTHOMAS/KPConv |
| Description | Fully convolutional network operating directly on points |
| Reference | @article{thomas2019KPConv, Author = {Thomas, Hugues and Qi, Charles R. and Deschaud, Jean-Emmanuel and Marcotegui, Beatriz and Goulette, Fran{c{c}}ois and Guibas, Leonidas J.}, Title = {KPConv: Flexible and Deformable Convolution for Point Clouds}, Journal = {The IEEE International Conference on Computer Vision (ICCV)}, Year = {2019} } |
| Hardware | Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz, 32GB RAM, Nvidia GeForce GTX 1080 Ti |
| Used additional training data | 0 |
| Last submission | 2019-03-27 12:09:46 |
| Is opensource | 1 |
| Number of submissions | 13 |
| A_IoU | OA | [s] | IoU 1 | IoU 2 | IoU 3 | IoU 4 | IoU 5 | IoU 6 | IoU 7 | IoU 8 | |
| Result | 0.746 | 0.929 | 600.00 | 0.909 | 0.822 | 0.842 | 0.479 | 0.949 | 0.400 | 0.773 | 0.797 |
| Rank | 4 | 9 | 9 | 11 | 12 | 9 | 6 | 4 | 4 | 1 | 3 |
| L1' | L2' | L3' | L4' | L5' | L6' | L7' | L8' | |
| L1 | 14137358 | 84361 | 58 | 4560 | 56085 | 26731 | 5985 | 2371 |
| L2 | 1179735 | 9164870 | 3248 | 322988 | 18351 | 5449 | 38 | 67 |
| L3 | 3 | 611 | 4623355 | 7928 | 32779 | 10450 | 745 | 0 |
| L4 | 12618 | 357335 | 798448 | 1686916 | 90221 | 71419 | 4230 | 707 |
| L5 | 17971 | 1424 | 4037 | 39899 | 30725230 | 118075 | 2542 | 2028 |
| L6 | 15605 | 11990 | 7552 | 118253 | 1142425 | 1067583 | 12640 | 8497 |
| L7 | 1890 | 248 | 363 | 916 | 9383 | 17662 | 203132 | 245 |
| L8 | 2372 | 12 | 0 | 2457 | 117545 | 36783 | 2817 | 689956 |