CVPR 2025 accepted papers on Image generation

481 of the 2,871 papers accepted at CVPR 2025 (16.8%) are on Image generation, compared with 16.6% at CVPR 2024. They fall into 7 topics.

See CVPR 2025 in the interactive map

Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.

Topics and papers

Text-to-image diffusion · 119 papers

OmniGen: Unified Image Generation (56 citations)
Diffusion-4K: Ultra-High-Resolution Image Synthesis with Latent Diffusion Models (20 citations)
VideoDPO: Omni-Preference Alignment for Video Diffusion Generation (18 citations)
DesignDiffusion: High-Quality Text-to-Design Image Generation with Diffusion Models (16 citations)
SleeperMark: Towards Robust Watermark against Fine-Tuning Text-to-image Diffusion Models (15 citations)
SALAD: Skeleton-aware Latent Diffusion for Text-driven Motion Generation and Editing (13 citations)
Diffusion Self-Distillation for Zero-Shot Customized Image Generation (11 citations)
POSTA: A Go-to Framework for Customized Artistic Poster Generation (11 citations)
Aesthetic Post-Training Diffusion Models from Generic Preferences with Step-by-step Preference Optimization (10 citations)
ART: Anonymous Region Transformer for Variable Multi-Layer Transparent Image Generation (10 citations)
Imagine and Seek: Improving Composed Image Retrieval with an Imagined Proxy (9 citations)
Towards Better Alignment: Training Diffusion Models with Reinforcement Learning Against Sparse Rewards (9 citations)

and 107 more in the interactive map.

Autoregressive and latent generation · 103 papers

Reconstruction vs. Generation: Taming Optimization Dilemma in Latent Diffusion Models (53 citations)
SIDA: Social Media Image Deepfake Detection, Localization and Explanation with Large Multimodal Model (53 citations)
MINIMA: Modality Invariant Image Matching (44 citations)
Teaching Large Language Models to Regress Accurate Image Quality Scores Using Score Distribution (44 citations)
Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis (38 citations)
3DTopia-XL: Scaling High-quality 3D Asset Generation via Primitive Diffusion (27 citations)
Dual-Interrelated Diffusion Model for Few-Shot Anomaly Image Generation (26 citations)
A Bias-Free Training Paradigm for More General AI-generated Image Detection (24 citations)
RoboTwin: Dual-Arm Robot Benchmark with Generative Digital Twins (24 citations)
Any-Resolution AI-Generated Image Detection by Spectral Learning (23 citations)
Unseen Visual Anomaly Generation (21 citations)
DexGrasp Anything: Towards Universal Robotic Dexterous Grasping with Physics Awareness (20 citations)

and 91 more in the interactive map.

Editing and style transfer · 78 papers

AnyEdit: Mastering Unified High-Quality Image Editing for Any Idea (28 citations)
Attention Distillation: A Unified Approach to Visual Characteristics Transfer (28 citations)
StableAnimator: High-Quality Identity-Preserving Human Image Animation (25 citations)
Stable Flow: Vital Layers for Training-Free Image Editing (21 citations)
K-LoRA: Unlocking Training-Free Fusion of Any Subject and Style LoRAs (20 citations)
SaMam: Style-aware State Space Model for Arbitrary Image Style Transfer (17 citations)
StyleStudio: Text-Driven Style Transfer with Selective Control of Style Elements (17 citations)
EchoMimicV2: Towards Striking, Simplified, and Semi-Body Human Animation (14 citations)
BooW-VTON: Boosting In-the-Wild Virtual Try-On via Mask-Free Pseudo Data Training (13 citations)
OSDFace: One-Step Diffusion Model for Face Restoration (13 citations)
VLOGGER: Multimodal Diffusion for Embodied Avatar Synthesis (13 citations)
Learning Flow Fields in Attention for Controllable Person Image Generation (12 citations)

and 66 more in the interactive map.

Fast diffusion sampling · 77 papers

DiffusionDrive: Truncated Diffusion Model for End-to-End Autonomous Driving (87 citations)
GoalFlow: Goal-Driven Flow Matching for Multimodal Trajectories Generation in End-to-End Autonomous Driving (30 citations)
Improving Diffusion Inverse Problem Solving with Decoupled Noise Annealing (20 citations)
Effective Cloud Removal for Remote Sensing Images by an Improved Mean-Reverting Denoising Model with Elucidated Design Space (17 citations)
Efficient Diffusion as Low Light Enhancer (17 citations)
Noise-Consistent Siamese-Diffusion for Medical Image Synthesis and Segmentation (16 citations)
Q-DiT: Accurate Post-Training Quantization for Diffusion Transformers (16 citations)
RAD: Region-Aware Diffusion Models for Image Inpainting (14 citations)
CleanDIFT: Diffusion Features without Noise (13 citations)
Scaling Inference Time Compute for Diffusion Models (13 citations)
DKDM: Data-Free Knowledge Distillation for Diffusion Models with Any Architecture (12 citations)
Flowing from Words to Pixels: A Noise-Free Framework for Cross-Modality Evolution (11 citations)

and 65 more in the interactive map.

Image restoration · 72 papers

HVI: A New Color Space for Low-light Image Enhancement (212 citations)
MambaIRv2: Attentive State Space Restoration (140 citations)
DarkIR: Robust Low-Light Image Restoration (96 citations)
MaIR: A Locality- and Continuity-Preserving Mamba for Image Restoration (59 citations)
Efficient Visual State Space Model for Image Deblurring (57 citations)
HyperFree: A Channel-adaptive and Tuning-free Foundation Model for Hyperspectral Remote Sensing Imagery (49 citations)
Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing (39 citations)
Iterative Predictor-Critic Code Decoding for Real-World Image Dehazing (36 citations)
CoA: Towards Real Image Dehazing via Compression-and-Adaptation (28 citations)
Learned Image Compression with Dictionary-based Entropy Model (23 citations)
Tokenize Image Patches: Global Context Fusion for Effective Haze Removal in Large Images (23 citations)
UltraFusion: Ultra High Dynamic Imaging using Exposure Fusion (23 citations)

and 60 more in the interactive map.

Super-resolution · 28 papers

CATANet: Efficient Content-Aware Token Aggregation for Lightweight Image Super-Resolution (62 citations)
Progressive Focused Transformer for Single Image Super-Resolution (47 citations)
Arbitrary-steps Image Super-resolution via Diffusion Inversion (45 citations)
Pixel-level and Semantic-level Adjustable Super-resolution: A Dual-LoRA Approach (34 citations)
DifIISR: A Diffusion Model with Gradient Guidance for Infrared Image Super-Resolution (30 citations)
Adversarial Diffusion Compression for Real-World Image Super-Resolution (26 citations)
TSD-SR: One-Step Diffusion with Target Score Distillation for Real-World Image Super-Resolution (26 citations)
FaithDiff: Unleashing Diffusion Priors for Faithful Image Super-resolution (25 citations)
HIIF: Hierarchical Encoding based Implicit Image Function for Continuous Super-resolution (18 citations)
Uncertainty-guided Perturbation for Image Super-Resolution Diffusion Model (18 citations)
DiffFNO: Diffusion Fourier Neural Operator (17 citations)
Latent Space Super-Resolution for Higher-Resolution Image Generation with Diffusion Models (11 citations)

and 16 more in the interactive map.