ImageNet-64

ImageNet-64 is a benchmark dataset containing 64x64 pixel images used to evaluate the performance of generative models, particularly in the context of Generative Adversarial Networks (GANs).

Papers using ImageNet-64 (10)

Variational Trajectory Optimization of Anisotropic Diffusion Schedules2026UCD: Unconditional Discriminator Promotes Nash Equilibrium in GANs2025Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator2025Generative Modeling with Bayesian Sample Inference2025Elucidating the Design Space of Diffusion-Based Generative Models2022 · 308 citesFast Sampling of Diffusion Models via Operator Learning2022 · 17 citesTRACT: Denoising Diffusion Models with Transitive Closure Time-Distillation2023 · 12 citesMultistep Consistency Models2024 · 3 citesTraining-free Linear Image Inverses via Flows2023EM Distillation for One-step Diffusion Models2024