Image Generation On Ffhq 256 X 256

Metrics

FID

Results

Performance results of various models on this benchmark

Model Name	FID	Paper Title	Repository
GANFormer	7.42	Generative Adversarial Transformers
BigGAN	11.48	A U-Net Based Discriminator for Generative Adversarial Networks
CLR-GAN	3.37	CLR-GAN: Improving GANs Stability and Quality via Consistent Latent Representation and Reconstruction
StyleNAT	2.05	StyleNAT: Giving Each Head a New Perspective
StyleGAN2-ada (Exposing)	5.30	Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
StyleGAN-XL (DINOv2)	-	StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets
LDM	8.11	-	-
INR-GAN-bil	4.95	Adversarial Generation of Continuous Images
Anycost GAN	3.35	Anycost GANs for Interactive Image Synthesis and Editing
R3GAN	2.75	The GAN is dead; long live the GAN! A Modern GAN Baseline
LeCAM (StyleGAN2 + ADA)	3.49	Regularizing Generative Adversarial Networks under Limited Data
Efficient-VDVAE	34.88	Efficient-VDVAE: Less is more
LDM (Exposing)	8.11	Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
Unleashing Transformers (DINOv2)	-	Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codes
InsGen (DINOv2)	-	Data-Efficient Instance Generation from Instance Discrimination
HiT-S	3.06	Improved Transformer for High-Resolution GANs
HiT-L	2.58	Improved Transformer for High-Resolution GANs
Very Deep VAE	-	Very Deep VAEs Generalize Autoregressive Models and Can Outperform Them on Images
StyleSwin	2.81	StyleSwin: Transformer-based GAN for High-resolution Image Generation
InsGen (Exposing)	3.46	Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models

0 of 51 row(s) selected.