Point Cloud Classification on ModelNet40 16-shot
89.79AccuracyCLIP2Point
Evaluation results
| Method | Links | |
|---|---|---|
| CLIP2PointEncoder=ViT-B/32, Pre-training=with2022.10 | 89.79 | |
| CLIP2PointEncoder=ViT-B/32, Pre-training=without2022.10 | 87.46 | |
| PointCLIPEncoder=ResNet101, Pre-training=without2022.10 | 87.2 | |
| CrossPointEncoder=DGCNN, Pre-training=with2022.10 | 84.48 | |
| Point-MAEEncoder=Transformer, Pre-training=with2022.10 | 84.2 | |
| PointCLIPEncoder=ViT-B/32, Pre-training=without2022.10 | 83.83 | |
| CrossPointEncoder=DGCNN, Pre-training=without2022.10 | 81.56 | |
| Point-MAEEncoder=Transformer, Pre-training=without2022.10 | 79.7 |