SAN: Inducing Metrizability of GAN with Discriminative Normalized Linear Layer | ✓ Link | 1.68 | | | | | | | StyleSAN-XL | 2023-01-30 |
StyleNAT: Giving Each Head a New Perspective | ✓ Link | 2.05 | | | | | | | StyleNAT | 2022-11-10 |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 2.11 | 229.42 | 0.79 | 0.41 | | 0.77 | 0.71 | StyleNAT (Exposing) | 2023-06-07 |
StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets | ✓ Link | 2.19 | | | | | | | StyleGAN-XL | 2022-02-01 |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 2.26 | 240.07 | 0.77 | 0.43 | | 0.68 | 0.63 | StyleGAN-XL (Exposing) | 2023-06-07 |
Compensation Sampling for Improved Convergence in Diffusion Models | ✓ Link | 2.57 | | | | | | | PDM+CS | 2023-12-11 |
Improved Transformer for High-Resolution GANs | ✓ Link | 2.58 | | | | | | | HiT-L | 2021-06-14 |
Polynomial Implicit Neural Representations For Large Diverse Datasets | ✓ Link | 2.72 | | | | | | | Poly-INR | 2023-03-20 |
The GAN is dead; long live the GAN! A Modern GAN Baseline | ✓ Link | 2.75 | | | | | | | R3GAN | 2025-01-09 |
StyleSwin: Transformer-based GAN for High-resolution Image Generation | ✓ Link | 2.81 | | | | | | | StyleSwin | 2021-12-20 |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 2.89 | 303.21 | 0.79 | 0.28 | | 0.71 | 0.64 | StyleSwin (Exposing) | 2023-06-07 |
Improved Transformer for High-Resolution GANs | ✓ Link | 2.95 | | | | | | | HiT-B | 2021-06-14 |
Improved Transformer for High-Resolution GANs | ✓ Link | 3.06 | | | | | | | HiT-S | 2021-06-14 |
Data-Efficient Instance Generation from Instance Discrimination | ✓ Link | 3.31 | | | | | | | InsGen | 2021-06-08 |
Anycost GANs for Interactive Image Synthesis and Editing | ✓ Link | 3.35 | | | | | | | Anycost GAN | 2021-03-04 |
CLR-GAN: Improving GANs Stability and Quality via Consistent Latent Representation and Reconstruction | ✓ Link | 3.37 | | 0.71 | 0.44 | | | | CLR-GAN | 2024-09-30 |
Projected GANs Converge Faster | ✓ Link | 3.39 | | 0.654 | 0.464 | | | | Projected-GAN (Official) | 2021-11-01 |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 3.46 | 436.26 | 0.64 | 0.13 | | | 0.51 | InsGen (Exposing) | 2023-06-07 |
Regularizing Generative Adversarial Networks under Limited Data | ✓ Link | 3.49 | | | | | | | LeCAM (StyleGAN2 + ADA) | 2021-04-07 |
Training Generative Adversarial Networks with Limited Data | ✓ Link | 3.62 | | | | | | | StyleGAN2 + ADA | 2020-06-11 |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 4.29 | 589.20 | 0.57 | 0.07 | | 0.31 | 0.30 | Projected-GAN (Exposing) | 2023-06-07 |
Image Generators with Conditionally-Independent Pixel Synthesis | ✓ Link | 4.38 | | | | | | | CIPS | 2020-11-27 |
Flow Matching in Latent Space | ✓ Link | 4.55 | | | | | | | LFM | 2023-07-17 |
Adversarial Generation of Continuous Images | ✓ Link | 4.95 | | | | | | | INR-GAN-bil | 2020-11-24 |
SWAGAN: A Style-based Wavelet-driven Generative Model | ✓ Link | 5.22 | | | | | | | SWAGAN-Bi | 2021-02-11 |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 5.30 | 514.78 | 0.59 | 0.06 | | 0.36 | 0.39 | StyleGAN2-ada (Exposing) | 2023-06-07 |
Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation | ✓ Link | 5.54 | | | | | | | UDM (RVE) + ST | 2021-06-10 |
Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codes | ✓ Link | 6.11 | | | | | | | Unleashing Transformers | 2021-11-24 |
Generative Adversarial Transformers | ✓ Link | 7.42 | | | | | | | GANFormer | 2021-03-01 |
A U-Net Based Discriminator for Generative Adversarial Networks | ✓ Link | 7.48 | | | | | | | U-Net GAN | 2020-02-28 |
Compositional Transformers for Scene Generation | ✓ Link | 7.77 | | | | | | | GANFormer2 | 2021-12-01 |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 8.11 | 226.72 | 0.81 | 0.44 | | 0.83 | 0.74 | LDM (Exposing) | 2023-06-07 |
[]() | | 8.11 | | | | | | | LDM | |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 9.02 | 393.45 | 0.76 | 0.24 | | 0.61 | 0.53 | Unleash-Trans (Exposing) | 2023-06-07 |
Taming Transformers for High-Resolution Image Synthesis | ✓ Link | 9.6 | | | | | | | VQGAN+Transformer | 2020-12-17 |
A U-Net Based Discriminator for Generative Adversarial Networks | ✓ Link | 11.48 | | | | | | | BigGAN | 2020-02-28 |
Gotta Go Fast When Generating Data with Score-Based Models | ✓ Link | 15.67 | | | | | | | VE (erel=0.01) | 2021-05-28 |
Gotta Go Fast When Generating Data with Score-Based Models | ✓ Link | 15.67 | | | | | | | VE (erel=0.02) | 2021-05-28 |
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models | ✓ Link | 34.88 | 514.16 | 0.86 | 0.14 | | 1.04 | 0.54 | Efficient-vdVAE (Exposing) | 2023-06-07 |
Efficient-VDVAE: Less is more | ✓ Link | 34.88 | | | | 0.53 | | | Efficient-VDVAE | 2022-03-25 |
[]() | | | 112.40 | 0.92 | 0.76 | | | | LDM (DINOv2) | |
StyleNAT: Giving Each Head a New Perspective | ✓ Link | | 229.72 | 0.79 | 0.41 | | | | StyleNAT (DINOv2) | 2022-11-10 |
StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets | ✓ Link | | 240.07 | 0.77 | 0.43 | | | | StyleGAN-XL (DINOv2) | 2022-02-01 |
StyleSwin: Transformer-based GAN for High-resolution Image Generation | ✓ Link | | 300.18 | 0.79 | 0.28 | | | | StyleSwin (DINOv2) | 2021-12-20 |
Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codes | ✓ Link | | 393.45 | 0.76 | 0.24 | | | | Unleashing Transformers (DINOv2) | 2021-11-24 |
Data-Efficient Instance Generation from Instance Discrimination | ✓ Link | | 436.26 | 0.64 | 0.13 | | | | InsGen (DINOv2) | 2021-06-08 |
Efficient-VDVAE: Less is more | ✓ Link | | 514.16 | 0.86 | 0.14 | | | | Efficient-VDVAE (DINOv2) | 2022-03-25 |
Training Generative Adversarial Networks with Limited Data | ✓ Link | | 514.78 | 0.59 | 0.06 | | | | StyleGAN2 + ADA (DINOv2) | 2020-06-11 |
[]() | | | 592.26 | 0.57 | 0.07 | | | | Projected-GAN (DINOv2) | |
Very Deep VAEs Generalize Autoregressive Models and Can Outperform Them on Images | ✓ Link | | | | | 0.61 | | | Very Deep VAE | 2020-11-20 |
NVAE: A Deep Hierarchical Variational Autoencoder | ✓ Link | | | | | 0.69 | | | NVAE w/ flow | 2020-07-08 |