3D Object Recognition on ModelNet40 8k points (test)
95AccuracyRECON++-L
Evaluation results
| Method | Links | |
|---|---|---|
| RECON++-LTraining Paradigm=Self-Supervised, Post-pretraining=true, Voting=true2024.02 | 95 | |
| PointGPT-LTraining Paradigm=Self-Supervised, Post-pretraining=true, Voting=true2024.02 | 94.9 | |
| RECON++-BTraining Paradigm=Self-Supervised, Post-pretraining=true, Voting=true2024.02 | 94.8 | |
| RECONTraining Paradigm=Self-Supervised, Post-pretraining=false, Voting=true2024.02 | 94.7 | |
| PointGPT-BTraining Paradigm=Self-Supervised, Post-pretraining=true, Voting=true2024.02 | 94.6 | |
| VPPTraining Paradigm=Self-Supervised, Post-pretraining=false, Voting=true2024.02 | 94.3 | |
| Point-MAETraining Paradigm=Self-Supervised, Post-pretraining=false, Voting=true2024.02 | 94 | |
| ACTTraining Paradigm=Self-Supervised, Post-pretraining=false, Voting=true2024.02 | 94 | |
| Point-BERTTraining Paradigm=Self-Supervised, Post-pretraining=false, Voting=true2024.02 | 93.8 | |
| PointNet++Training Paradigm=Supervised, Post-pretraining=false, Voting=true2024.02 | 91.9 | |
| PointNetTraining Paradigm=Supervised, Post-pretraining=false, Voting=true2024.02 | 90.8 |