UNCSN++ (RVE) + ST

Reported on 4 benchmarks across 1 task · 1 paper · 3 SOTA

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

Medical4 results

Image GenerationonSTL-10
FID· 2021-06-10
7.71
best: 6.91 (Diffusion ProjectedGAN)
SOTA
Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation arXiv:2106.05527
Image GenerationonSTL-10
Inception score· 2021-06-10
13.43
SOTA
Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation arXiv:2106.05527
Image GenerationonCelebA-HQ 256x256
FID· 2021-06-10
7.16
best: 3.15 (RDM)
SOTA
Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation arXiv:2106.05527
Image GenerationonCelebA 64x64
bits/dimension· 2021-06-10
1.97
best: 1.83 (Efficient-VDVAE)
Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation arXiv:2106.05527