NCP-VAE

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

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

Medical3 results

Image GenerationonCelebA 64x64
FID· 2020-10-06
5.25
best: 1.27 (DDPM-IP)
SOTA
A Contrastive Learning Approach for Training Variational Autoencoder Priors arXiv:2010.02917
Image GenerationonCelebA 256x256
FID· 2020-10-06
24.79
best: 3.25 (StyleSwin)
A Contrastive Learning Approach for Training Variational Autoencoder Priors arXiv:2010.02917
Image GenerationonCIFAR-10
FID· 2020-10-06
24.08
best: 1.22 (GMem)
A Contrastive Learning Approach for Training Variational Autoencoder Priors arXiv:2010.02917