Lepard: Learning partial point cloud matching in rigid and deformable scenes

Open paper
Benchmark Model Rank Results
partial-point-cloud-matching-on-4dmatchLi and Harada (θc=0.05)#1NFMR: 83.9IR: 80.9
partial-point-cloud-matching-on-4dmatchLi and Harada (θc=0.1)#2NFMR: 83.7IR: 82.7
partial-point-cloud-matching-on-4dmatchLi and Harada (θc=0.2)#3NFMR: 82.2IR: 85.4
partial-point-cloud-matching-on-4dmatchPredator (5000)#4NFMR: 56.8IR: 59.3
partial-point-cloud-matching-on-4dmatchPredator (3000)#5NFMR: 56.4IR: 60.4
partial-point-cloud-matching-on-4dmatchD3Feat (5000)#6NFMR: 56.1IR: 55.3
partial-point-cloud-matching-on-4dmatchD3Feat (3000)#7NFMR: 55.5IR: 54.7
partial-point-cloud-matching-on-4dmatchPredator (1000)#8NFMR: 53.3IR: 60
partial-point-cloud-matching-on-4dmatchD3Feat (1000)#9NFMR: 51.6IR: 52.7

Rank counts only results with a code link.