Image Generation On Lsun Churches 256 X 256

FID

Résultats

Résultats de performance de divers modèles sur ce benchmark

Tableau comparatif

Nom du modèle	FID
unleashing-transformers-parallel-token	4.07
wavelet-diffusion-models-are-fast-and	5.06
tackling-the-generative-learning-trilemma-1	5.25
adversarial-generation-of-continuous-images	4.04
diffusion-gan-training-gans-with-diffusion	1.85
swagan-a-style-based-wavelet-driven	4.97
truncated-diffusion-probabilistic-models	3.98
bellman-optimal-step-size-straightening-of	-
flow-matching-in-latent-space	5.54
Modèle 10	2.66
a-style-based-generator-architecture-for	4.21
denoising-diffusion-probabilistic-models	7.89
diffusion-gan-training-gans-with-diffusion	3.17
pseudo-numerical-methods-for-diffusion-models-1	8.69
gotta-go-fast-when-generating-data-with-score	26.46
clr-gan-improving-gans-stability-and-quality	3.43
ensembling-off-the-shelf-models-for-gan	1.72
analyzing-and-improving-the-image-quality-of	3.86
polarity-sampling-quality-and-diversity	3.92
image-generators-with-conditionally	2.92
stylenat-giving-each-head-a-new-perspective	3.4
transgan-two-transformers-can-make-one-strong	8.94
styleswin-transformer-based-gan-for-high-1	2.95
projected-gans-converge-faster	1.59
gotta-go-fast-when-generating-data-with-score	26.46
progressive-growing-of-gans-for-improved	6.42
msg-gan-multi-scale-gradients-gan-for-more	5.2