Unconditional 3D Point Cloud Generation on ShapeNet Airplane v1 (test)
0.2205MMD (CD)PSF
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| PSFSampling steps (N)=12022.12 | 0.2205 | 0.3661 | 46.17 | 52.59 | |
| PointFlow2022.12 | 0.2243 | 0.3901 | 47.9 | 46.41 | |
| PVD2022.12 | 0.2243 | 0.3803 | 48.88 | 52.09 | |
| SoftFlow2022.12 | 0.2309 | 0.3745 | 46.91 | 47.9 | |
| PVD-DDIMSampling protocol=DDIM, Sampling steps (N)=1002022.12 | 0.2434 | 0.3991 | 44.23 | 49.75 | |
| DPF-Net2022.12 | 0.2642 | 0.4086 | 46.17 | 48.89 | |
| l-GAN (CD)Training metric=CD2022.12 | 0.3398 | 0.5832 | 38.52 | 21.23 | |
| Shape-GF2022.12 | 2.703 | 0.6592 | 40.74 | 40.49 |