ICML 2025 accepted papers on Image generation
237 of the 3,342 papers accepted at ICML 2025 (7.1%) are on Image generation, compared with 5.2% at ICML 2024. They fall into 7 topics.
See ICML 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
Fast diffusion sampling · 103 papers
- PepTune: De Novo Generation of Therapeutic Peptides with Multi-Objective-Guided Discrete Diffusion (10 citations)
- A General Framework for Inference-time Scaling and Steering of Diffusion Models (5 citations)
- An analytic theory of creativity in convolutional diffusion models (4 citations)
- Normalizing Flows are Capable Generative Models (4 citations)
- Synthetic Face Datasets Generation via Latent Space Exploration from Brownian Identity Diffusion (4 citations)
- FireFlow: Fast Inversion of Rectified Flow for Image Semantic Editing (2 citations)
- Is Noise Conditioning Necessary for Denoising Generative Models? (2 citations)
- A Mixture-Based Framework for Guiding Diffusion Models (1 citation)
- Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching (1 citation)
- Compressed Image Generation with Denoising Diffusion Codebook Models (1 citation)
- Distributional Diffusion Models with Scoring Rules (1 citation)
- Feynman-Kac Correctors in Diffusion: Annealing, Guidance, and Product of Experts (1 citation)
and 91 more in the interactive map.
Autoregressive and latent generation · 72 papers
- All-atom Diffusion Transformers: Unified generative modelling of molecules and materials (7 citations)
- Diffusion on Language Model Encodings for Protein Sequence Generation (7 citations)
- WyckoffDiff -- A Generative Diffusion Model for Crystal Symmetry (3 citations)
- Efficient Fine-Grained Guidance for Diffusion Model Based Symbolic Music Generation (1 citation)
- Efficient Generative Modeling with Residual Vector Quantization-Based Tokens (1 citation)
- Geometric Representation Condition Improves Equivariant Molecule Generation (1 citation)
- Kinetic Langevin Diffusion for Crystalline Materials Generation (1 citation)
- Latent Diffusion Planning for Imitation Learning (1 citation)
- SongGen: A Single Stage Auto-regressive Transformer for Text-to-Song Generation (1 citation)
- Towards Black-Box Membership Inference Attack for Diffusion Models (1 citation)
- UniMate: A Unified Model for Mechanical Metamaterial Generation, Property Prediction, and Condition Confirmation (1 citation)
- WMAdapter: Adding WaterMark Control to Latent Diffusion Models (1 citation)
and 60 more in the interactive map.
Text-to-image diffusion · 38 papers
- RealRAG: Retrieval-augmented Realistic Image Generation via Self-reflective Contrastive Learning (1 citation)
- AdvI2I: Adversarial Image Attack on Image-to-Image Diffusion Models
- Angle Domain Guidance: Latent Diffusion Requires Rotation Rather Than Extrapolation
- CHATS: Combining Human-Aligned Optimization and Test-Time Sampling for Text-to-Image Generation
- Control and Realism: Best of Both Worlds in Layout-to-Image without Training
- D-Fusion: Direct Preference Optimization for Aligning Diffusion Models with Visually Consistent Samples
- David and Goliath: Small One-step Model Beats Large Diffusion with Score Post-training
- Diffuse Everything: Multimodal Diffusion Models on Arbitrary State Spaces
- DreamDPO: Aligning Text-to-3D Generation with Human Preferences via Direct Preference Optimization
- EraseAnything: Enabling Concept Erasure in Rectified Flow Transformers
- Event-Customized Image Generation
- FlexControl: Computation-Aware Conditional Control with Differentiable Router for Text-to-Image Generation
and 26 more in the interactive map.
Editing and style transfer · 11 papers
- Taming Rectified Flow for Inversion and Editing (2 citations)
- LightningDrag: Lightning Fast and Accurate Drag-based Image Editing Emerging from Videos (1 citation)
- Balancing Preservation and Modification: A Region and Semantic Aware Metric for Instruction-Based Image Editing
- Diffusion Counterfactual Generation with Semantic Abduction
- DragLoRA: Online Optimization of LoRA Adapters for Drag-based Image Editing in Diffusion Model
- EasyRef: Omni-Generalized Group Image Reference for Diffusion Models via Multimodal LLM
- Editable Noise Map Inversion: Encoding Target-image into Noise For High-Fidelity Image Manipulation
- Highly Compressed Tokenizer Can Generate Without Training
- Resolving Lexical Bias in Model Editing
- Rethinking Score Distilling Sampling for 3D Editing and Generation
- Enhancing Visual Localization with Cross-Domain Image Generation
Image restoration · 10 papers
- FourierMamba: Fourier Learning Integration with State Space Models for Image Deraining (2 citations)
- ArrayDPS: Unsupervised Blind Speech Separation with a Diffusion Prior
- DEALing with Image Reconstruction: Deep Attentive Least Squares
- Human Body Restoration with One-Step Diffusion Model and A New Benchmark
- IRBridge: Solving Image Restoration Bridge with Pre-trained Generative Diffusion Models
- LotteryCodec: Searching the Implicit Representation in a Random Network for Low-Complexity Image Compression
- Stochastic Deep Restoration Priors for Imaging Inverse Problems
- Feature out! Let Raw Image as Your Condition for Blind Face Restoration
- Spherical-Nested Diffusion Model for Panoramic Image Outpainting
- Ultra Lowrate Image Compression with Semantic Residual Coding and Compression-aware Diffusion