NeurIPS 2026 accepted papers on Image generation
456 of the 6,080 papers accepted at NeurIPS 2026 (7.5%) are on Image generation, compared with 7.6% at NeurIPS 2025. They fall into 7 topics.
See NeurIPS 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 · 209 papers
- A Mean-Field Framework for Inference-Time Distributional Control of Diffusion Models
- A Steerable Deep Network for Model-Free Diffusion MRI Registration
- A Theoretical Analysis of Why Masked Diffusion Models Mitigate the Reversal Curse
- ABC: Any-Subset Autoregression via Non-Markovian Diffusion Bridges in Continuous Time and Space
- Accelerating Diffusion Language Models via Structured Suffix Modeling
- Accelerating Rectified Flow Models via Trajectory-Aware Caching
- Amortized Guidance for Image Inpainting with Pretrained Diffusion Models
- Analyzing and Guiding Zero-Shot Posterior Sampling in Diffusion Models
- Asymmetric Flow Models
- Attention-Based Sampler for Diffusion Language Models
- Attention-Discounted Adaptive Sampler for Masked Diffusion Language Models
- Backbone-Equated Diffusion OOD via Sparse Internal Snapshots
and 197 more in the interactive map.
Autoregressive and latent generation · 99 papers
- AHPA: Adaptive Hierarchical Prior Alignment for Diffusion Transformers
- AMGenC: Generating Charge Balanced Amorphous Materials
- AudioCALM: Continuous Autoregressive Language Modeling for Universal Audio Generation
- Balancing Image Compression and Generation with Bootstrapped Tokenization
- BIAS-ID: A Framework for Analyzing Transformation Biases in AI-Generated Image Detectors
- BitDance: Scaling Autoregressive Generative Models with Binary Tokens
- BSTabDiff: Block-Subunit Diffusion Priors for High-Dimensional Tabular Data Generation
- Channel-wise Vector Quantization
- Coarse-to-Fine Compositional Diffusion for Long-Horizon Planning
- Continuous Latent Diffusion Language Model
- DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion Convergence
- DecQ: Detail-Condensing Queries for Enhanced Reconstruction and Generation in Representation Autoencoders
and 87 more in the interactive map.
Text-to-image diffusion · 74 papers
- Text-to-CAD Evaluation with CADTests (1 citation)
- A Matched-Budget Audit Framework for Recaptioned Image-Text Supervision Distributions
- CASL: Concept-Aligned Sparse Latents for Interpreting Diffusion Models
- ColorConceptBench: A Benchmark for Probabilistic Color-Concept Understanding in Text-to-Image Models
- Consistency-Preserving Concept Erasure via Unsafe–Safe Pairing and Directional Fisher-weighted Adaptation
- Controllable Road Marking Generation
- Diff-Aid: Inference-time Adaptive Interaction Denoising for Rectified Text-to-Image Generation
- Diff-Instruct with Diffused Reward: Towards Principled One-step Generator RL
- DifFRACT: Diffusion Feature Reconstruction and Attribution for Circuit Tracing
- Diffusion-DRF: Free, Rich, and Differentiable Reward for Video Diffusion Fine-Tuning
- DynT2I-Eval: A Dynamic Evaluation Framework for Text-to-Image Models
- EDITORS Know Your Style! Editing LoRA Subspaces for Stylistic Attribution and Imitation
and 62 more in the interactive map.
Image restoration · 30 papers
- A Principled Self-Referenced Early Stopping Approach for Deep Image Prior
- AnyBokeh: Physics-Guided Any-to-Any Bokeh Editing with Optical Fingerprint Transfer
- Bricker to BRACE: A Bracket Exposure RAW Dataset and Restoration Model for Flicker-Banding
- Continuous Expert Assembly: Instance-Conditioned Low-Rank Residuals for All-in-One Image Restoration
- Inference-Time Search Using Side Information for Diffusion-Based Image Reconstruction
- IRPO: Boosting Image Restoration via Post-training GRPO
- Physics Unrolled Neural Operator for Wireless Field Modeling
- PRISM: Rethinking Atmospheric Scattering Reconstruction as a Unified Understanding and Restoration Model for Real-world Dehazing
- RIFLE: Removal of Image Flicker-Banding via Latent Diffusion Enhancement
- SP$^3$: Spherical Priors for Plug-and-Play Restoration
- A General Filter-Enhanced Approach to Smartphone Hyperspectral Imaging
- Bound-Conditioned Latent Inference for Progressive Image Compression
and 18 more in the interactive map.
Editing and style transfer · 25 papers
- CLeaR: A Unified Framework for Resolving the Leakage–Degradation Dilemma in Style Transfer
- Do Image Editing Models Understand Lighting?
- Edit the Bits, Diff the Codes: Bitwise Residual Editing for Visual Autoregressive Models
- Edit-R2: Context-Aware Reinforcement Learning for Multi-Turn Image Editing
- Non-Colliding Biometric Identities for Digital Entities: Geometry, Capacity, and Million-Scale Virtual Identity Provisioning
- Steering Fields: Adaptive Vector Fields for Safe Image Generation and Beyond
- Text-Vision Co-Instructed Image Editing
- Why Do DiT Editors Drift? Plug-and-Play Low Frequency Alignment in VAE Latent Space
- Anatomy-Preserving Unpaired Medical Image Translation via Shared Latent Anchoring
- Beyond Unidirectional: Unsupervised Trajectory Learning for Omnidirectional Controllable Underwater Image Enhancement
- Can We Model the Artifacts Explicitly? Disentangle Artifacts via Pairwise Edit Relations for Image Manipulation Localization
- Constrained Look-ahead Guidance for Interference-Aware Flow Editing
and 13 more in the interactive map.
Super-resolution · 17 papers
- DegBins: Degradation-Driven Binning for Depth Super-Resolution
- EditBridge: Towards Faithful and Efficient Ultra-High-Resolution Image Editing
- Everything at Every Scale: Scale-Invariant Diffusion with Continuous Super-Resolution
- FluxFlow: Conservative Flow-Matching for Astronomical Image Super-Resolution
- LucidNFT: LR-Anchored Multi-Reward Preference Optimization for Flow-Based Real-World Super-Resolution
- NSARM: Next-Scale Autoregressive Modeling for Robust Real-World Image Super-Resolution
- PRISM: Prior Rectification and Uncertainty-Aware Structure Modeling for Diffusion-Based Text Image Super-Resolution
- When Latents Forget Pixels: Restoring Fidelity in Diffusion Transformer Super-Resolution
- CoVisIT: Cross-modal Prior Guided Diffusion Model for Visible-to-Infrared Image Translation
- Decoupling Direction and Magnitude: Language-Steered Flow Matching for Super-Resolution in the Dark
- Epigenomics-Guided Flow Matching for 3D Genome Super-Resolution
- Hardware-Friendly Token-Group Activation Quantization for Low-Bit Mamba Super-Resolution
and 5 more in the interactive map.