Point Cloud Registration on ModelNet40 twice-sampled (TS) unseen categories (test)
2.079RMSE (Rotation)OMNet
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| OMNet2021.03 | 2.079 | 0.619 | 0.0077 | 0.0042 | 1.241 | 0.0154 | |
| OMNet2021.03 | 2.514 | 1.004 | 0.0177 | 0.0078 | 1.949 | 0.0154 | |
| PRNet2021.03 | 2.712 | 1.372 | 0.0171 | 0.0081 | 2.607 | 0.0237 | |
| OMNetPartial manner=RPMNet, Noise=Gaussian noise, Data variant=twice-sampled (TS)2021.03 | 4.356 | 1.924 | 0.0486 | 0.0223 | 3.834 | 0.0476 | |
| FMR2021.03 | 5.041 | 2.304 | 0.0383 | 0.0158 | 4.525 | 0.0314 | |
| FMR2021.03 | 5.119 | 2.349 | 0.0296 | 0.0147 | 4.553 | 0.0292 | |
| RPMNetPartial manner=RPMNet, Noise=Gaussian noise, Data variant=twice-sampled (TS)2021.03 | 6.16 | 2.407 | 0.0618 | 0.0274 | 4.913 | 0.0589 | |
| FGR2021.03 | 6.39 | 1.24 | 0.0375 | 0.0081 | 2.204 | 0.0156 | |
| PRNet2021.03 | 6.506 | 3.472 | 0.0388 | 0.0257 | 6.789 | 0.051 | |
| Symmetric ICP2021.03 | 6.599 | 5.962 | 0.0654 | 0.0592 | 11.713 | 0.1134 | |
| DCP2021.03 | 6.754 | 4.366 | 0.0612 | 0.0403 | 8.566 | 0.0807 | |
| IDAM2021.03 | 6.852 | 1.761 | 0.054 | 0.0138 | 3.433 | 0.0275 | |
| Symmetric ICP2021.03 | 7.415 | 6.552 | 0.0759 | 0.0684 | 13.113 | 0.1315 | |
| DCP2021.03 | 7.683 | 4.747 | 0.0675 | 0.0427 | 9.764 | 0.0862 | |
| IDAM2021.03 | 8.346 | 4.54 | 0.059 | 0.0329 | 8.679 | 0.0656 | |
| FMRPartial manner=RPMNet, Noise=Gaussian noise, Data variant=twice-sampled (TS)2021.03 | 11.674 | 7.4 | 0.1364 | 0.0867 | 14.121 | 0.187 | |
| Symmetric ICPPartial manner=RPMNet, Noise=Gaussian noise, Data variant=twice-sampled (TS)2021.03 | 12.576 | 10.987 | 0.1478 | 0.1203 | 21.807 | 0.256 | |
| DCPPartial manner=RPMNet, Noise=Gaussian noise, Data variant=twice-sampled (TS)2021.03 | 12.73 | 9.556 | 0.1072 | 0.0774 | 12.173 | 0.1586 | |
| Go-ICP2021.03 | 13.572 | 3.416 | 0.0448 | 0.0152 | 6.688 | 0.0299 | |
| Go-ICP2021.03 | 14.162 | 4.19 | 0.0533 | 0.0206 | 8.286 | 0.0409 | |
| ICP2021.03 | 17.236 | 8.61 | 0.0817 | 0.0434 | 16.824 | 0.0855 | |
| PointNetLK2021.03 | 18.294 | 9.73 | 0.0917 | 0.0526 | 18.845 | 0.1042 | |
| ICP2021.03 | 18.458 | 9.335 | 0.0915 | 0.0505 | 18.194 | 0.0993 | |
| DeepGMR2021.03 | 18.89 | 9.322 | 0.087 | 0.0559 | 17.513 | 0.1108 | |
| PointNetLK2021.03 | 21.041 | 10.74 | 0.113 | 0.0629 | 20.438 | 0.125 | |
| ICPPartial manner=RPMNet, Noise=Gaussian noise, Data variant=twice-sampled (TS)2021.03 | 21.893 | 13.402 | 0.1963 | 0.1278 | 26.632 | 0.2679 | |
| DeepGMR2021.03 | 23.472 | 12.863 | 0.0987 | 0.0658 | 24.425 | 0.1298 | |
| PointNetLKPartial manner=RPMNet, Noise=Gaussian noise, Data variant=twice-sampled (TS)2021.03 | 29.733 | 21.154 | 0.267 | 0.1937 | 42.027 | 0.3964 | |
| FGR2021.03 | 29.838 | 16.361 | 0.147 | 0.0818 | 31.153 | 0.163 | |
| FGRPartial manner=RPMNet, Noise=Gaussian noise, Data variant=twice-sampled (TS)2021.03 | 46.213 | 30.116 | 0.3034 | 0.2141 | 58.968 | 0.4364 |