ICML 2026 accepted papers on Image generation
429 of the 6,341 papers accepted at ICML 2026 (6.8%) are on Image generation, compared with 7.1% at ICML 2025. They fall into 7 topics.
See ICML 2026 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
Fast diffusion sampling · 211 papers
- One-step Latent-free Image Generation with Pixel Mean Flows (1 citation)
- A theory of learning data statistics in diffusion models, from easy to hard
- Active Tabular Augmentation via Policy-Guided Diffusion Inpainting
- Adversarial Flow Models
- Alleviating Sparse Rewards by Modeling Step-Wise and Long-Term Sampling Effects in Flow-Based GRPO
- Ambient Dataloops: Generative Models for Dataset Refinement
- Anomaly-Preference Image Generation
- Are First-Order Diffusion Samplers Really Slower? A Fast Forward-Value Approach
- AugMask: Training Diffusion Models on Incomplete Tabular Data via Stochastic Augmentation and Masking
- Balancing Understanding and Generation in Discrete Diffusion Models
- Bayesian Tensor Decomposition with Diffusion Model Prior
- Beyond Confidence: Adaptive and Coherent Decoding for Diffusion Language Models
and 199 more in the interactive map.
Autoregressive and latent generation · 91 papers
- Scalable Single-Cell Gene Expression Generation with Latent Diffusion Models (2 citations)
- Multimodal Latent Language Modeling with Next-Token Diffusion (1 citation)
- SPATIA: Multimodal Generation and Prediction of Spatial Cell Phenotypes (1 citation)
- (1D) Ordered Tokens Enable Efficient Test-Time Search
- ArcVQ-VAE: A Spherical Vector Quantization Framework with ArcCosine Additive Margin
- Autoregressive Image Generation with Masked Bit Modeling
- Both Semantics and Reconstruction Matter: Making Representation Encoders Ready for Text-to-Image Generation and Editing
- Coevolutionary Continuous Discrete Diffusion: Make Your Diffusion Language Model a Latent Reasoner
- Compositional Generative Modeling from Decentralized Data
- Conformal Reliability: A New Evaluation Metric for Conditional Generation
- Cross-Chirality Generalization by Axial Vectors for Hetero-Chiral Protein-Peptide Interaction Design
- CSD: Content-aware Speculative Decoding for Efficient Image Generation
and 79 more in the interactive map.
Text-to-image diffusion · 58 papers
- A Unified Framework for Diffusion Model Unlearning with f-Divergence
- Alignment-Guided Score Matching for Text-to-Image Alignment in Diffusion Models
- Attention Sinks in Diffusion Transformers: A Causal Analysis
- Breaking the Lock-in: Diversifying Text-to-Image Generation via Representation Modulation
- Causal-Adapter: Taming Text-to-Image Diffusion for Faithful Counterfactual Generation
- Cert-LAS: Toward Certified Model Ownership Verification for Text-to-Image Diffusion Models via Layer-Adaptive Smoothing
- Concept Removal for Frontier Image Generative Models
- Concept Removal Guidance: Evidence-Calibrated Negative Guidance for Safe Diffusion Sampling
- Diagnosing and Correcting Concept Omission in Multimodal Diffusion Transformers
- Equalized Generative Treatment: Matching f-divergences for Fairness in Generative Models
- Erased but Not Forgotten: How Backdoors Compromise Concept Erasure
- Exposing Hidden Biases in Text-to-Image Models via Automated Prompt Search
and 46 more in the interactive map.
Image restoration · 29 papers
- AnyBand-Diff: A Unified Remote Sensing Image Generation and Band Repair Framework with Spectral Priors
- Bend the Basics: Degradation-Aware Deformable Tokenization for All-in-One Image Restoration
- CoD-Lite: Real-Time Diffusion-Based Generative Image Compression
- Degradation-Aware Metric Prompting for Hyperspectral Image Restoration
- Dual-View Predictive Diffusion: Lightweight Speech Enhancement via Spectrogram-Image Synergy
- Hyperspectral Image Fusion with Spectral-Band and Fusion-Scale Agnosticism
- Image Restoration via Diffusion Models with Dynamic Resolution
- Learning to Refine: Spectral-Decoupled Iterative Refinement Framework for Precipitation Nowcasting
- Light Up Your Face: A Physically Consistent Dataset and Diffusion Model for Face Fill-Light Enhancement
- Normalization-equivariant Diffusion Models: Learning Posterior Samplers From Noisy And Partial Measurements
- Plug-and-Play Diffusion Meets ADMM: Dual-Variable Coupling for Robust Medical Image Reconstruction
- Resolving Blind Inverse Problems under Dynamic Range Compression via Structured Forward Operator Modeling
and 17 more in the interactive map.
Editing and style transfer · 28 papers
- AesFormer: Transform Everyday Photos into Beautiful Memories
- Balancing Fidelity and Diversity in Diffusion Models via Symmetric Attention Decomposition: Hopfield Perspective
- CoCoEdit: Content-Consistent Image Editing via Region Regularized Reinforcement Learning
- Content-Style Identification via Differential Independence
- DirectEdit: Step-Level Accurate Inversion for Flow-Based Image Editing
- DLEBench: Evaluating Small-scale Object Editing Ability for Instruction-based Image Editing Model
- From Statics to Dynamics: Physics-Aware Image Editing with Latent Transition Priors
- Geometrically Constrained Stenosis Editing in Coronary Angiography via Entropic Optimal Transport
- OpenGPT-4o-Image: A Comprehensive Dataset for Advanced Image Generation and Editing
- Orthogonal Concept Erasure for Diffusion Models
- PhotoAgent: Exploratory Visual Aesthetic Planning with Large Vision Models
- Polyphonia: Zero-Shot Timbre Transfer in Polyphonic Music with Acoustic-Informed Attention Calibration
and 16 more in the interactive map.
Super-resolution · 10 papers
- Coloring the Noise: Adversarial Sobolev Alignment for Faithful Image Super Resolution
- Hierarchical Image Tokenization for Multi-Scale Image Super Resolution
- Joint Geometric and Trajectory Consistency Learning for One-Step Real-World Super-Resolution
- LiteVSR: Lightweight Adaptation of Frozen Diffusion Transformers for Video Super-Resolution
- PODiff: Latent Diffusion in Proper Orthogonal Decomposition Space for Scientific Super-Resolution
- Q-DiT4SR: Exploration of Detail-Preserving Diffusion Transformer Quantization for Real-World Image Super-Resolution
- Enhanced Latent-Space Adversarial Training for Super-Resolution
- One-Step Residual Shifting Diffusion for Image Super-Resolution via Distillation
- SSDCN: Spatial-Spectral Dual-Clustering-based Network for Hyperspectral Image Super-resolution
- WEVSR: Video Diffusion Generators for Real-World Video Super‑Resolution with Wavelet-Enhanced VAE Encoder