ECCV 2024 accepted papers on Image generation
379 of the 2,387 papers accepted at ECCV 2024 (15.9%) are on Image generation, compared with 11.6% at ECCV 2022. They fall into 7 topics.
See ECCV 2024 in the interactive map
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
Text-to-image diffusion · 146 papers
- PixArt-Sigma: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image Generation (322 citations)
- ZipLoRA: Any Subject in Any Style by Effectively Merging LoRAs (149 citations)
- To Generate or Not? Safety-Driven Unlearned Diffusion Models Are Still Easy To Generate Unsafe Images ... For Now (125 citations)
- Investigating Style Similarity in Diffusion Models (124 citations)
- Concept Sliders: LoRA Adaptors for Precise Control in Diffusion Models (118 citations)
- Implicit Style-Content Separation using B-LoRA (104 citations)
- TextDiffuser-2: Unleashing the Power of Language Models for Text Rendering (99 citations)
- MobileDiffusion: Instant Text-to-Image Generation on Mobile Devices (98 citations)
- ControlNet++: Improving Conditional Controls with Efficient Consistency Feedback (97 citations)
- DOCCI: Descriptions of Connected and Contrasting Images (70 citations)
- Receler: Reliable Concept Erasing of Text-to-Image Diffusion Models via Lightweight Erasers (70 citations)
- CogView3: Finer and Faster Text-to-Image Generation via Relay Diffusion (69 citations)
and 134 more in the interactive map.
Image restoration · 60 papers
- MambaIR: A Simple Baseline for Image Restoration with State-Space Model (865 citations)
- DiffBIR: Toward Blind Image Restoration with Generative Diffusion Prior (322 citations)
- InstructIR: High-Quality Image Restoration Following Human Instructions (145 citations)
- AutoDIR: Automatic All-in-One Image Restoration with Latent Diffusion (99 citations)
- When Fast Fourier Transform Meets Transformer for Image Restoration (99 citations)
- OneRestore: A Universal Restoration Framework for Composite Degradation (92 citations)
- Efficient Frequency-Domain Image Deraining with Contrastive Regularization (86 citations)
- A Comparative Study of Image Restoration Networks for General Backbone Network Design (74 citations)
- Fast Context-Based Low-Light Image Enhancement via Neural Implicit Representations (55 citations)
- WeConvene: Learned Image Compression with Wavelet-Domain Convolution and Entropy Model (37 citations)
- Seeing the Unseen: A Frequency Prompt Guided Transformer for Image Restoration (32 citations)
- ZeST: Zero-Shot Material Transfer from a Single Image (30 citations)
and 48 more in the interactive map.
Fast diffusion sampling · 57 papers
- Adversarial Diffusion Distillation (633 citations)
- SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers (372 citations)
- BrushNet: A Plug-and-Play Image Inpainting Model with Decomposed Dual-Branch Diffusion (172 citations)
- "BK-SDM: A Lightweight, Fast, and Cheap Version of Stable Diffusion" (65 citations)
- Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidance (52 citations)
- DGInStyle: Domain-Generalizable Semantic Segmentation with Image Diffusion Models and Stylized Semantic Control (48 citations)
- Distilling Diffusion Models into Conditional GANs (48 citations)
- FMBoost: Boosting Latent Diffusion with Flow Matching (33 citations)
- Lossy Image Compression with Foundation Diffusion Models (31 citations)
- MixDQ: Memory-Efficient Few-Step Text-to-Image Diffusion Models with Metric-Decoupled Mixed Precision Quantization (28 citations)
- Do text-free diffusion models learn discriminative visual representations? (27 citations)
- Diffusion for Natural Image Matting (26 citations)
and 45 more in the interactive map.
Autoregressive and latent generation · 41 papers
- ZigMa: A DiT-style Zigzag Mamba Diffusion Model (167 citations)
- DiffiT: Diffusion Vision Transformers for Image Generation (102 citations)
- GLARE: Low Light Image Enhancement via Generative Latent Feature based Codebook Retrieval (77 citations)
- Leveraging Representations from Intermediate Encoder-blocks for Synthetic Image Detection (62 citations)
- RealGen: Retrieval Augmented Generation for Controllable Traffic Scenarios (48 citations)
- GIVT: Generative Infinite-Vocabulary Transformers (46 citations)
- Safe-Sim: Safety-Critical Closed-Loop Traffic Simulation with Diffusion-Controllable Adversaries (46 citations)
- LaWa: Using Latent Space for In-Generation Image Watermarking (44 citations)
- A Watermark-Conditioned Diffusion Model for IP Protection (43 citations)
- Zero-Shot Detection of AI-Generated Images (37 citations)
- HiDiffusion: Unlocking Higher-Resolution Creativity and Efficiency in Pretrained Diffusion Models (35 citations)
- HERGen: Elevating Radiology Report Generation with Longitudinal Data (33 citations)
and 29 more in the interactive map.
Editing and style transfer · 38 papers
- LightenDiffusion: Unsupervised Low-Light Image Enhancement with Latent-Retinex Diffusion Models (176 citations)
- Improving Diffusion Models for Authentic Virtual Try-on in the Wild (171 citations)
- A Task is Worth One Word: Learning with Task Prompts for High-Quality Versatile Image Inpainting (146 citations)
- Arc2Face: A Foundation Model for ID-Consistent Human Faces (106 citations)
- ReNoise: Real Image Inversion Through Iterative Noising (84 citations)
- Face Adapter for Pre-Trained Diffusion Models with Fine-Grained ID and Attribute Control (60 citations)
- ObjectDrop: Bootstrapping Counterfactuals for Photorealistic Object Removal and Insertion (57 citations)
- Wear-Any-Way: Manipulable Virtual Try-on via Sparse Correspondence Alignment (37 citations)
- DiffFAS: Face Anti-Spoofing via Generative Diffusion Models (29 citations)
- EditShield: Protecting Unauthorized Image Editing by Instruction-guided Diffusion Models (29 citations)
- StableDrag: Stable Dragging for Point-based Image Editing (25 citations)
- SwapAnything: Enabling Arbitrary Object Swapping in Personalized Image Editing (25 citations)
and 26 more in the interactive map.
Super-resolution · 28 papers
- Pixel-Aware Stable Diffusion for Realistic Image Super-Resolution and Personalized Stylization (159 citations)
- SMFANet: A Lightweight Self-Modulation Feature Aggregation Network for Efficient Image Super-Resolution (127 citations)
- HiT-SR: Hierarchical Transformer for Efficient Image Super-Resolution (83 citations)
- MTKD: Multi-Teacher Knowledge Distillation for Image Super-Resolution (42 citations)
- Contourlet Residual for Prompt Learning Enhanced Infrared Image Super-Resolution (32 citations)
- XPSR: Cross-modal Priors for Diffusion-based Image Super-Resolution (29 citations)
- SPIRE: Semantic Prompt-Driven Image Restoration (26 citations)
- You Only Need One Step: Fast Super-Resolution with Stable Diffusion via Scale Distillation (16 citations)
- Inf-DiT: Upsampling any-resolution image with memory-efficient diffusion transformer. (13 citations)
- UCIP: A Universal Framework for Compressed Image Super-Resolution using Dynamic Prompt (13 citations)
- Brain-ID: Learning Contrast-agnostic Anatomical Representations for Brain Imaging (12 citations)
- Accelerating Image Super-Resolution Networks with Pixel-Level Classification (10 citations)
and 16 more in the interactive map.
GANs · 9 papers
- A Closer Look at GAN Priors: Exploiting Intermediate Features for Enhanced Model Inversion Attacks (28 citations)
- Blind Image Deconvolution by Generative-based Kernel Prior and Initializer via Latent Encoding (7 citations)
- A high-quality robust diffusion framework for corrupted dataset (5 citations)
- Linearly Controllable GAN: Unsupervised Feature Categorization and Decomposition for Image Generation and Manipulation (4 citations)
- Nickel and Diming Your GAN: A Dual-Method Approach to Enhancing GAN Efficiency via Knowledge Distillation (3 citations)
- 2S-ODIS: Two-Stage Omni-Directional Image Synthesis by Geometric Distortion Correction (2 citations)
- CLR-GAN: Improving GANs Stability and Quality via Consistent Latent Representation and Reconstruction (2 citations)
- Rejection Sampling IMLE: Designing Priors for Better Few-Shot Image Synthesis (2 citations)
- Efficient Training with Denoised Neural Weights (1 citation)
