ICLR 2022 accepted papers on Image generation
46 of the 1,095 papers accepted at ICLR 2022 (4.2%) are on Image generation, compared with 4.7% at ICLR 2021. They fall into 6 topics.
See ICLR 2022 in the interactive map
Submitting to ICLR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Fast diffusion sampling · 15 papers
- Progressive Distillation for Fast Sampling of Diffusion Models (186 citations)
- Tackling the Generative Learning Trilemma with Denoising Diffusion GANs (141 citations)
- Pseudo Numerical Methods for Diffusion Models on Manifolds (124 citations)
- Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models (74 citations)
- Label-Efficient Semantic Segmentation with Diffusion Models (28 citations)
- Step-unrolled Denoising Autoencoders for Text Generation (28 citations)
- BDDM: Bilateral Denoising Diffusion Models for Fast and High-Quality Speech Synthesis (26 citations)
- Diffusion-Based Voice Conversion with Fast Maximum Likelihood Sampling Scheme (25 citations)
- PriorGrad: Improving Conditional Denoising Diffusion Models with Data-Dependent Adaptive Prior (23 citations)
- Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality (22 citations)
- Score-Based Generative Modeling with Critically-Damped Langevin Diffusion (7 citations)
- Denoising Likelihood Score Matching for Conditional Score-based Data Generation (6 citations)
and 3 more in the interactive map.
Autoregressive and latent generation · 11 papers
- Vector-quantized Image Modeling with Improved VQGAN (91 citations)
- InfinityGAN: Towards Infinite-Pixel Image Synthesis (20 citations)
- Responsible Disclosure of Generative Models Using Scalable Fingerprinting (20 citations)
- Learning Super-Features for Image Retrieval (13 citations)
- Conditional Image Generation by Conditioning Variational Auto-Encoders (10 citations)
- Generative Models as a Data Source for Multiview Representation Learning (6 citations)
- On the relation between statistical learning and perceptual distances (5 citations)
- ClimateGAN: Raising Climate Change Awareness by Generating Images of Floods (2 citations)
- Generalized rectifier wavelet covariance models for texture synthesis (1 citation)
- Prototype memory and attention mechanisms for few shot image generation
- Learning Disentangled Representation by Exploiting Pretrained Generative Models: A Contrastive Learning View
GANs · 8 papers
- Finding an Unsupervised Image Segmenter in each of your Deep Generative Models (20 citations)
- Collapse by Conditioning: Training Class-conditional GANs with Limited Data (16 citations)
- Mind the Gap: Domain Gap Control for Single Shot Domain Adaptation for Generative Adversarial Networks (16 citations)
- Chunked Autoregressive GAN for Conditional Waveform Synthesis (6 citations)
- GATSBI: Generative Adversarial Training for Simulation-Based Inference (5 citations)
- MaGNET: Uniform Sampling from Deep Generative Network Manifolds Without Retraining (2 citations)
- Pretraining Text Encoders with Adversarial Mixture of Training Signal Generators (2 citations)
- Stein Latent Optimization for Generative Adversarial Networks
Image restoration · 6 papers
- Solving Inverse Problems in Medical Imaging with Score-Based Generative Models (132 citations)
- Clean Images are Hard to Reblur: Exploiting the Ill-Posed Inverse Task for Dynamic Scene Deblurring (8 citations)
- Interpretable Unsupervised Diversity Denoising and Artefact Removal (7 citations)
- Generative Modeling with Optimal Transport Maps (5 citations)
- The Spectral Bias of Polynomial Neural Networks
- Fixed Neural Network Steganography: Train the images, not the network
Editing and style transfer · 4 papers
- SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations (301 citations)
- StyleAlign: Analysis and Applications of Aligned StyleGAN Models (19 citations)
- Retriever: Learning Content-Style Representation as a Token-Level Bipartite Graph (4 citations)
- Distilling GANs with Style-Mixed Triplets for X2I Translation with Limited Data
