CVPR 2021 accepted papers on Image generation
196 of the 1,661 papers accepted at CVPR 2021 (11.8%) are on Image generation, compared with 11.4% at CVPR 2020. They fall into 7 topics.
See CVPR 2021 in the interactive map
Submitting to CVPR? 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
Image restoration · 64 papers
- Multi-Stage Progressive Image Restoration (2,086 citations)
- Contrastive Learning for Compact Single Image Dehazing (940 citations)
- Towards Real-World Blind Face Restoration With Generative Facial Prior (525 citations)
- Neighbor2Neighbor: Self-Supervised Denoising From Single Noisy Images (446 citations)
- See Through Gradients: Image Batch Recovery via GradInversion (438 citations)
- PSD: Principled Synthetic-to-Real Dehazing Guided by Physical Priors (375 citations)
- GAN Prior Embedded Network for Blind Face Restoration in the Wild (292 citations)
- Large-Capacity Image Steganography Based on Invertible Neural Networks (277 citations)
- Ultra-High-Definition Image Dehazing via Multi-Guided Bilateral Learning (260 citations)
- PD-GAN: Probabilistic Diverse GAN for Image Inpainting (238 citations)
- NBNet: Noise Basis Learning for Image Denoising With Subspace Projection (235 citations)
- Learning Multi-Scale Photo Exposure Correction (229 citations)
and 52 more in the interactive map.
Editing and style transfer · 50 papers
- Encoding in Style: A StyleGAN Encoder for Image-to-Image Translation (1,083 citations)
- StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation (380 citations)
- TediGAN: Text-Guided Diverse Face Image Generation and Manipulation (360 citations)
- VITON-HD: High-Resolution Virtual Try-On via Misalignment-Aware Normalization (304 citations)
- ArtFlow: Unbiased Image Style Transfer via Reversible Neural Flows (254 citations)
- Parser-Free Virtual Try-On via Distilling Appearance Flows (210 citations)
- One Shot Face Swapping on Megapixels (175 citations)
- Progressive Semantic-Aware Style Transformation for Blind Face Restoration (174 citations)
- PISE: Person Image Synthesis and Editing With Decoupled GAN (151 citations)
- Image-to-Image Translation via Hierarchical Style Disentanglement (146 citations)
- DG-Font: Deformable Generative Networks for Unsupervised Font Generation (138 citations)
- CoSMo: Content-Style Modulation for Image Retrieval With Text Feedback (134 citations)
and 38 more in the interactive map.
GANs · 28 papers
- Closed-Form Factorization of Latent Semantics in GANs (489 citations)
- House-GAN++: Generative Adversarial Layout Refinement Network towards Intelligent Computational Agent for Professional Architects (150 citations)
- HistoGAN: Controlling Colors of GAN-Generated and Real Images via Color Histograms (146 citations)
- Exploiting Spatial Dimensions of Latent in GAN for Real-Time Image Editing (128 citations)
- Adversarial Generation of Continuous Images (124 citations)
- DivCo: Diverse Conditional Image Synthesis via Contrastive Generative Adversarial Network (90 citations)
- Anycost GANs for Interactive Image Synthesis and Editing (81 citations)
- Generative Hierarchical Features From Synthesizing Images (80 citations)
- Positional Encoding As Spatial Inductive Bias in GANs (71 citations)
- Navigating the GAN Parameter Space for Semantic Image Editing (66 citations)
- Teachers Do More Than Teach: Compressing Image-to-Image Models (53 citations)
- Content-Aware GAN Compression (50 citations)
and 16 more in the interactive map.
Super-resolution · 27 papers
- Unsupervised Degradation Representation Learning for Blind Super-Resolution (418 citations)
- Interpreting Super-Resolution Networks With Local Attribution Maps (346 citations)
- Exploring Sparsity in Image Super-Resolution for Efficient Inference (312 citations)
- GLEAN: Generative Latent Bank for Large-Factor Image Super-Resolution (268 citations)
- ClassSR: A General Framework to Accelerate Super-Resolution Networks by Data Characteristic (209 citations)
- MASA-SR: Matching Acceleration and Spatial Adaptation for Reference-Based Image Super-Resolution (176 citations)
- Flow-Based Kernel Prior With Application to Blind Super-Resolution (164 citations)
- Unsupervised Real-World Image Super Resolution via Domain-Distance Aware Training (161 citations)
- Scene Text Telescope: Text-Focused Scene Image Super-Resolution (159 citations)
- Deep Burst Super-Resolution (147 citations)
- MR Image Super-Resolution With Squeeze and Excitation Reasoning Attention Network (117 citations)
- AdderSR: Towards Energy Efficient Image Super-Resolution (98 citations)
and 15 more in the interactive map.
Autoregressive and latent generation · 24 papers
- Taming Transformers for High-Resolution Image Synthesis (2,386 citations)
- Pre-Trained Image Processing Transformer (2,102 citations)
- Generating Diverse Structure for Image Inpainting With Hierarchical VQ-VAE (230 citations)
- Real-Time High-Resolution Background Matting (229 citations)
- VIGOR: Cross-View Image Geo-Localization Beyond One-to-One Retrieval (226 citations)
- Leveraging Line-Point Consistence To Preserve Structures for Wide Parallax Image Stitching (152 citations)
- On Feature Normalization and Data Augmentation (142 citations)
- Image Generators With Conditionally-Independent Pixel Synthesis (129 citations)
- Variational Transformer Networks for Layout Generation (124 citations)
- Mask Guided Matting via Progressive Refinement Network (121 citations)
- Image Inpainting Guided by Coherence Priors of Semantics and Textures (119 citations)
- Im2Vec: Synthesizing Vector Graphics Without Vector Supervision (95 citations)
and 12 more in the interactive map.
Fast diffusion sampling · 2 papers
- Soft-IntroVAE: Analyzing and Improving the Introspective Variational Autoencoder (45 citations)
- Synthesize-It-Classifier: Learning a Generative Classifier Through Recurrent Self-Analysis (1 citation)
Text-to-image diffusion · 1 paper
- Cross-Modal Contrastive Learning for Text-to-Image Generation (333 citations)
