OcCo

Reported on 5 benchmarks across 4 tasks · 1 paper · 4 SOTA

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

Computer Vision5 results

Shape Representation Of 3D Point CloudsonScanObjectNN 10-way (10-shot)
Overall Accuracy· 2020-10-02
57
best: 69.6 (CrossMoCo)
SOTA
Unsupervised Point Cloud Pre-Training via Occlusion Completion arXiv:2010.01089
3D Point Cloud ClassificationonScanObjectNN 10-way (10-shot)
Overall Accuracy· 2020-10-02
57
best: 69.6 (CrossMoCo)
SOTA
Unsupervised Point Cloud Pre-Training via Occlusion Completion arXiv:2010.01089
3D Point Cloud Linear ClassificationonScanObjectNN
Overall Accuracy· 2020-10-02
78.3
best: 86.06 (CrossMoCo)
SOTA
Unsupervised Point Cloud Pre-Training via Occlusion Completion arXiv:2010.01089
3D Point Cloud ReconstructiononScanObjectNN 10-way (10-shot)
Overall Accuracy· 2020-10-02
57
best: 69.6 (CrossMoCo)
SOTA
Unsupervised Point Cloud Pre-Training via Occlusion Completion arXiv:2010.01089
3D Point Cloud Linear ClassificationonModelNet40
Overall Accuracy· 2020-10-02
89.2
best: 93.6 (ReCon++)
Unsupervised Point Cloud Pre-Training via Occlusion Completion arXiv:2010.01089