ImageNet 256×256
Emerging26papers using it
2021first seen
'ImageNet 256×256' is a dataset used to evaluate the performance of image synthesis models, specifically in terms of their ability to generate high-quality images at a resolution of 256 by 256 pixels.
Papers using ImageNet 256×256 (26)
- Mean Flows for One-step Generative ModelingEfficient Generative Modeling with Residual Vector Quantization-Based TokensDiffusion Image Generation with Explicit Modeling of Data Manifold GeometryWhat Matters for Diffusion-Friendly Latent Manifold? Prior-Aligned Autoencoders for Latent DiffusionLaminating Representation Autoencoders for Efficient DiffusionGroup Diffusion: Enhancing Image Generation by Unlocking Cross-Sample CollaborationTerminal Velocity MatchingLatent Denoising Makes Good TokenizersUnified Continuous Generative ModelsLearning to Integrate Diffusion ODEs by Averaging the DerivativesREPA-E: Unlocking VAE for End-to-End Tuning with Latent Diffusion TransformersInductive Moment MatchingDiffusion Models Beat GANs on Image SynthesisScalable Diffusion Models with TransformersSiT: Exploring Flow and Diffusion-based Generative Models with Scalable
Interpolant TransformersPatch Diffusion: Faster and More Data-Efficient Training of Diffusion
ModelsRefining Generative Process with Discriminator Guidance in Score-based
Diffusion ModelsOn Distillation of Guided Diffusion ModelsFast Training of Diffusion Models with Masked TransformersAll are Worth Words: A ViT Backbone for Diffusion ModelsSAN: Inducing Metrizability of GAN with Discriminative Normalized Linear
LayerEfficient Diffusion Training via Min-SNR Weighting StrategyDiffFit: Unlocking Transferability of Large Diffusion Models via Simple
Parameter-Efficient Fine-TuningRelay Diffusion: Unifying diffusion process across resolutions for image
synthesisThink While You Generate: Discrete Diffusion with Planned DenoisingMaskBit: Embedding-free Image Generation via Bit Tokens