TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/Exhaustive Grid Search

Exhaustive Grid Search

Reported on 62 benchmarks across 2 tasks

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision62 results

  • Point Cloud RegistrationonFP-R-E
    RRE (degrees)
    0.005
    best: 2.29 (GeoTransformer)
  • Point Cloud RegistrationonFP-R-E
    RTE (cm)
    0.002
    best: 1.57 (GeoTransformer)
  • Point Cloud RegistrationonFP-R-E
    Recall (3cm, 10 degrees)
    99.64
    best: 99.76 (GeDi)
  • Point Cloud RegistrationonFP-R-H
    RRE (degrees)
    0.01
    best: 1.77 (FCGF+SC2-PCR )
  • Point Cloud RegistrationonFP-R-H
    RTE (cm)
    0.007
    best: 1.69 (GeoTransformer)
  • Point Cloud RegistrationonFP-R-H
    Recall (3cm, 10 degrees)
    78
    best: 99.41 (GeDi)
  • Point Cloud RegistrationonKITTI (trained on 3DMatch)
    Success Rate
    94.95
    best: 98.92 (GeDi)
  • Point Cloud RegistrationonFP-O-M
    RRE (degrees)
    0.03
    best: 2.45 (GeoTransformer)
  • Point Cloud RegistrationonFP-O-M
    RTE (cm)
    0.017
    best: 1.94 (GeoTransformer)
  • Point Cloud RegistrationonFP-O-M
    Recall (3cm, 10 degrees)
    88.06
  • Point Cloud Registrationon3DMatch (at least 30% overlapped - FCGF setting)
    Recall (0.3m, 15 degrees)
    84.11
    best: 95 (NgeNet)
  • Point Cloud RegistrationonFP-O-E
    RRE (degrees)
    0.067
    best: 2.3 (GeoTransformer)
  • Point Cloud RegistrationonFP-O-E
    RTE (cm)
    0.035
    best: 1.57 (GeoTransformer)
  • Point Cloud RegistrationonFP-O-E
    Recall (3cm, 10 degrees)
    99.47
    best: 99.88 (FPFH+SC2-PCR )
  • Point Cloud RegistrationonKITTI (FCGF setting)
    Recall (0.6m, 5 degrees)
    94.59
    best: 99.5 (GeoTransformer)
  • Point Cloud RegistrationonETH (trained on 3DMatch)
    Recall (30cm, 5 degrees)
    94.25
  • Point Cloud RegistrationonFP-T-M
    RRE (degrees)
    0.005
    best: 2.29 (GeoTransformer)
  • Point Cloud RegistrationonFP-T-M
    RTE (cm)
    0.002
    best: 1.58 (GeoTransformer)
  • Point Cloud RegistrationonFP-T-M
    Recall (3cm, 10 degrees)
    99.82
  • Point Cloud RegistrationonFP-T-E
    RRE (degrees)
    0.005
    best: 2.32 (GeoTransformer)
  • Point Cloud RegistrationonFP-T-E
    RTE (cm)
    0.002
    best: 1.59 (GeoTransformer)
  • Point Cloud RegistrationonFP-T-E
    Recall (3cm, 10 degrees)
    99.7
    best: 99.76 (FPFH+SC2-PCR)
  • Point Cloud RegistrationonFP-R-M
    RRE (degrees)
    0.003
    best: 2.26 (GeoTransformer)
  • Point Cloud RegistrationonFP-R-M
    RTE (cm)
    0.003
    best: 2.71 (GO-ICP)
  • Point Cloud RegistrationonFP-R-M
    Recall (3cm, 10 degrees)
    97.92
    best: 99.94 (GeDi)
  • Point Cloud RegistrationonFP-T-H
    RRE (degrees)
    0.005
    best: 2.27 (GeoTransformer)
  • Point Cloud RegistrationonFP-T-H
    RTE (cm)
    0.002
    best: 1.57 (GeoTransformer)
  • Point Cloud RegistrationonFP-T-H
    Recall (3cm, 10 degrees)
    98.81
    best: 99.7 (GeDi)
  • Point Cloud RegistrationonFP-O-H
    RRE (degrees)
    0.477
    best: 2.68 (FCGF+SC2-PCR)
  • Point Cloud RegistrationonFP-O-H
    RTE (cm)
    0.234
    best: 2.22 (GeoTransformer)
  • Point Cloud RegistrationonFP-O-H
    Recall (3cm, 10 degrees)
    37.06
    best: 38.85 (FPFH+SC2-PCR)
  • 3D Point Cloud InterpolationonFP-R-E
    RRE (degrees)
    0.005
    best: 2.29 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-R-E
    RTE (cm)
    0.002
    best: 1.57 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-R-E
    Recall (3cm, 10 degrees)
    99.64
    best: 99.76 (GeDi)
  • 3D Point Cloud InterpolationonFP-R-H
    RRE (degrees)
    0.01
    best: 1.77 (FCGF+SC2-PCR )
  • 3D Point Cloud InterpolationonFP-R-H
    RTE (cm)
    0.007
    best: 1.69 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-R-H
    Recall (3cm, 10 degrees)
    78
    best: 99.41 (GeDi)
  • 3D Point Cloud InterpolationonKITTI (trained on 3DMatch)
    Success Rate
    94.95
    best: 98.92 (GeDi)
  • 3D Point Cloud InterpolationonFP-O-M
    RRE (degrees)
    0.03
    best: 2.45 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-O-M
    RTE (cm)
    0.017
    best: 1.94 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-O-M
    Recall (3cm, 10 degrees)
    88.06
  • 3D Point Cloud Interpolationon3DMatch (at least 30% overlapped - FCGF setting)
    Recall (0.3m, 15 degrees)
    84.11
    best: 95 (NgeNet)
  • 3D Point Cloud InterpolationonFP-O-E
    RRE (degrees)
    0.067
    best: 2.3 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-O-E
    RTE (cm)
    0.035
    best: 1.57 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-O-E
    Recall (3cm, 10 degrees)
    99.47
    best: 99.88 (FPFH+SC2-PCR )
  • 3D Point Cloud InterpolationonKITTI (FCGF setting)
    Recall (0.6m, 5 degrees)
    94.59
    best: 99.5 (GeoTransformer)
  • 3D Point Cloud InterpolationonETH (trained on 3DMatch)
    Recall (30cm, 5 degrees)
    94.25
  • 3D Point Cloud InterpolationonFP-T-M
    RRE (degrees)
    0.005
    best: 2.29 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-T-M
    RTE (cm)
    0.002
    best: 1.58 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-T-M
    Recall (3cm, 10 degrees)
    99.82
  • 3D Point Cloud InterpolationonFP-T-E
    RRE (degrees)
    0.005
    best: 2.32 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-T-E
    RTE (cm)
    0.002
    best: 1.59 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-T-E
    Recall (3cm, 10 degrees)
    99.7
    best: 99.76 (FPFH+SC2-PCR)
  • 3D Point Cloud InterpolationonFP-R-M
    RRE (degrees)
    0.003
    best: 2.26 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-R-M
    RTE (cm)
    0.003
    best: 2.71 (GO-ICP)
  • 3D Point Cloud InterpolationonFP-R-M
    Recall (3cm, 10 degrees)
    97.92
    best: 99.94 (GeDi)
  • 3D Point Cloud InterpolationonFP-T-H
    RRE (degrees)
    0.005
    best: 2.27 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-T-H
    RTE (cm)
    0.002
    best: 1.57 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-T-H
    Recall (3cm, 10 degrees)
    98.81
    best: 99.7 (GeDi)
  • 3D Point Cloud InterpolationonFP-O-H
    RRE (degrees)
    0.477
    best: 2.68 (FCGF+SC2-PCR)
  • 3D Point Cloud InterpolationonFP-O-H
    RTE (cm)
    0.234
    best: 2.22 (GeoTransformer)
  • 3D Point Cloud InterpolationonFP-O-H
    Recall (3cm, 10 degrees)
    37.06
    best: 38.85 (FPFH+SC2-PCR)