D3Feat: Joint Learning of Dense Detection and Description of 3D Local Features

Benchmark Model Rank Results
point-cloud-registration-on-kittiD3Feat-pred#2Success Rate: 99.81
point-cloud-registration-on-3dmatch-trained-on-kittiD3Feat-pred#3Recall: 0.627
point-cloud-registration-on-3dmatch-benchmarkD3Feat-Pred#7Feature Matching Recall: 95.8
point-cloud-registration-on-3dmatch-benchmarkD3Feat-rand#8Feature Matching Recall: 95.3
point-cloud-registration-on-eth-trained-on-3dmatchD3Feat-pred#7Feature Matching Recall: 0.563
point-cloud-registration-on-3dmatch-at-least-30-overlapped-sample-5k-interest-pointsD3Feat (reported in PREDATOR)#10Recall ( correspondence RMSE below 0.2): 81.6
point-cloud-registration-on-3dlomatch-10-30-overlapD3Feat (reported in PREDATOR)#12Recall ( correspondence RMSE below 0.2): 37.2
point-cloud-registration-on-kitti-trained-on-3dmatchD3Feat-pred#12Success Rate: 36.76