AdaCrossNet

Reported on 7 benchmarks across 6 tasks

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 CloudsonModelNet40
Overall Accuracy
93.1
best: 95.3 (PointGST)
3D Point Cloud ClassificationonModelNet40
Overall Accuracy
93.1
best: 95.3 (PointGST)
3D Point Cloud Linear ClassificationonModelNet40
Overall Accuracy· uses extra data
91.8
best: 93.6 (ReCon++)
3D Point Cloud Linear ClassificationonScanObjectNN
Overall Accuracy
82.1
best: 86.06 (CrossMoCo)
3D Point Cloud ReconstructiononModelNet40
Overall Accuracy
93.1
best: 95.3 (PointGST)

Semantic SegmentationonShapeNet-Part
Class Average IoU
85.1
best: 87.7 (Feature Geometric Net (FG-Net))

10-shot image generationonShapeNet-Part
Class Average IoU
85.1
best: 87.7 (Feature Geometric Net (FG-Net))