awesome-text-to-image-studies

This GitHub repository summarizes papers and resources related to the text-to-image (T2I) generation task.

[!NOTE] This document serves as the homepage of the whole GitHub repo. Papers are summarized according to different research directions, published years, and conferences.

The topics section summarizes papers that are highly related to T2I generation according to different properties, e.g., prerequisites of T2I generation, diffusion models with other techniques (e.g., Diffusion Transformer, LLMs, Mamba, etc.), and diffusion models for other tasks.

If you have any suggestions about this reåpository, please feel free to start a new issue or pull requests.

Recent news of this GitHub repo are listed as follows.

🔥 [Dec. 11th, 2025] Our paper titled "StableV2V: Stablizing Shape Consistency in Video-to-Video Editing" is accepted at TCSVT 2025!

🔥 [Nov. 19th] We have released our latest paper titled "StableV2V: Stablizing Shape Consistency in Video-to-Video Editing", with the correponding code, model weights, and a testing benchmark DAVIS-Edit open-sourced. Feel free to check them out from the links!

• [Apr. 26th] Update a new topic: Diffusion Models Meet Federated Learning. See the topics section for more details!
• [Mar. 28th] The official AAAI 2024 paper list are released! Official version of PDFs and BibTeX references are updated accordingly.
• [Mar. 21th] The topics section has been updated. This section aims to offer paper lists that are summarized according to other properties of diffusion models, e.g., Diffusion Transformer-based methods, diffusion models for NLP, diffusion models integrated with LLMs, etc. The corresponding references of these papers are also concluded in reference.bib.
• [Mar. 7th] All available CVPR, ICLR, and AAAI 2024 papers and references are updated.
• [Mar. 1st] Websites of the off-the-shelf text-to-image generation products and toolkits are summarized.

Contents

• Products
• To-Do Lists
• Papers
  • Survey Papers
  • Text-to-Image Generation
    • Year 2025
    • Year 2024
    • Year 2023
    • Year 2022
    • Year 2021
    • Year 2020
  • Conditional Text-to-Image Generation
    • Year 2025
    • Year 2024
    • Year 2023
    • Year 2022
  • Personalized Text-to-Image Generation
    • Year 2025
    • Year 2024
    • Year 2023
  • Text-Guided Image Editing
    • Year 2026
    • Year 2025
    • Year 2024
    • Year 2023
    • Year 2022
  • Text Image Generation
    • Year 2024
• Datasets
• Toolkits
• Q&A
• References
• Star History
• WeChat Group

To-Do Lists

• Published Papers on Conferences
  • Update NeurIPS 2025 Papers
  • Update ICCV 2025 Papers
  • Update CVPR 2025 Papers
  • Update ICLR 2025 Papers
  • Update NeurIPS 2024 Papers
  • Update ECCV 2024 Papers
  • Update CVPR 2024 Papers
    • Update ⚠️ Papers and References
    • Update arXiv References into the Official Version
  • Update AAAI 2024 Papers
    • Update ⚠️ Papers and References
    • Update arXiv References into the Official Version
  • Update ICLR 2024 Papers
  • Update NeurIPS 2023 Papers
• Regular Maintenance of Preprint arXiv Papers and Missed Papers

<🎯Back to Top>

Products

Name	Year	Website	Specialties
AI Art Generator	2025	link	AI art generation platform with resolution presets, style transfer, upscaling, and video creation.
Nano Image Art	2025	link	Create stunning AI images — powered by Google's Nano Banana Pro for next-gen quality, smart editing, and intelligent prompts.
Fast Image AI	2025	link	Fast Image AI instantly transforms your photos into stunning styles like Ghibli, Sketch, and Pixar. Effortlessly control image elements and create amazing effects with just one click.
Gempix2 (Nano Banana 2)	2025	link	Free AI image generation platform with text-to-image, AI editing, and video generation support
Raphael AI	-	link	Free unlimited AI image generator (text-to-image, no signup)
Stable Diffusion 3	2024	link	Diffusion Transformer-based Stable Diffusion
Stable Video	2024	link	High-quality high-resolution images
DALL-E 3	2023	link	Collaborate with ChatGPT
Ideogram	2023	link	Text images
Playground	2023	link	Athestic images
HiDream.ai	2023	link	-
Dashtoon	2023	link	Text-to-Comic Generation
WHEE	2023	link	WHEE is an online AI generation tool, which can be applied for T2I generation, I2I generation, SR, inpainting, outpainting, image variation, virtural try-on, etc.
Vega AI	2023	link	Vega AI is an online AI generation tool, which can be applied for T2I generation, I2I generation, SR, T2V generation, I2V generation, etc.
Wujie AI	2022	link	The Chinese name is "无界AI", offering AIGC resources and online services
Midjourney	2022	link	Powerful close-sourced generation tool

<🎯Back to Top>

Papers

Survey Papers

• Text-to-Image Generation
  • Year 2024
    • ACM Computing Surveys
      • Diffusion Models: A Comprehensive Survey of Methods and Applications [Paper]
  • Year 2023
    • TPAMI
      • Diffusion Models in Vision: A Survey [Paper] [Code]
    • arXiv
      • Text-to-image Diffusion Models in Generative AI: A Survey [Paper]
      • State of the Art on Diffusion Models for Visual Computing [Paper]
  • Year 2022
    • arXiv
      • Efficient Diffusion Models for Vision: A Survey [Paper]
• Conditional Text-to-Image Generation
  • Year 2024
    • arXiv
      • Controllable Generation with Text-to-Image Diffusion Models: A Survey [Paper]
• Text-Guided Image Editing
  • Year 2024
    • arXiv
      • Diffusion Model-Based Image Editing: A Survey [Paper] [Code]

<🎯Back to Top>

Text-to-Image Generation

• Year 2026
  • ICLR
    • Asynchronous Denoising Diffusion Models for Aligning Text-to-Image Generation [Paper]
    • Beyond Text-to-Image: Liberating Generation with a Unified Discrete Diffusion Model [Paper]
    • Consistent Text-to-Image Generation via Scene De-Contextualization [Paper]
    • CoDi: Subject-Consistent and Pose-Diverse Text-to-Image Generation [Paper]
    • Continual Unlearning for Text-to-Image Diffusion Models: A Regularization Perspective [Paper]
    • Culture in Action: Evaluating Text-to-Image Models through Social Activities [Paper]
    • DeLeaker: Dynamic Inference-Time Reweighting For Semantic Leakage Mitigation in Text-to-Image Models [Paper]
    • Diverse Text-to-Image Generation via Contrastive Noise Optimization [Paper]
    • Easier Painting Than Thinking: Can Text-to-Image Models Set the Stage, but Not Direct the Play? [Paper]
    • Everything in Its Place: Benchmarking Spatial Intelligence of Text-to-Image Models [Paper]
    • Guidance Matters: Rethinking the Evaluation Pitfall for Text-to-Image Generation [Paper]
    • ImageDoctor: Diagnosing Text-to-Image Generation via Grounded Image Reasoning [Paper]
    • Interleaving Reasoning for Better Text-to-Image Generation [Paper]
    • Localized Concept Erasure in Text-to-Image Diffusion Models via High-Level Representation Misdirection [Paper]
    • Long-Text-to-Image Generation via Compositional Prompt Decomposition [Paper]
    • RePrompt: Reasoning-Augmented Reprompting for Text-to-Image Generation via Reinforcement Learning [Paper]
    • Sample Reward Soups: Query-efficient Multi-Reward Guidance for Text-to-Image Diffusion Models [Paper]
    • SenseFlow: Scaling Distribution Matching for Flow-based Text-to-Image Distillation [Paper]
    • TIPO: Text to Image with Text Pre-sampling for Prompt Optimization [Paper]
    • VisualPrompter: Semantic-Aware Prompt Optimization with Visual Feedback for Text-to-Image Synthesis [Paper]
    • From Broad Exploration to Stable Synthesis: Entropy-Guided Optimization for Autoregressive Image Generation [Paper]
    • From Prediction to Perfection: Introducing Refinement to Autoregressive Image Generation [Paper]
    • GIR-Bench: Versatile Benchmark for Generating Images with Reasoning [Paper]
    • Group Critical-token Policy Optimization for Autoregressive Image Generation [Paper]
    • ImagenWorld: Stress-Testing Image Generation Models with Explainable Human Evaluation on Open-ended Real-World Tasks [Paper]
    • Lavida-O: Elastic Large Masked Diffusion Models for Unified Multimodal Understanding and Generation [Paper]
    • LogiStory: A Logic-Aware Framework for Multi-Image Story Visualization [Paper]
    • MANZANO: A Simple and Scalable Unified Multimodal Model with a Hybrid Vision Tokenizer [Paper]
    • MILR: Improving Multimodal Image Generation via Test-Time Latent Reasoning [Paper]
    • MMaDA-Parallel: Multimodal Large Diffusion Language Models for Thinking-Aware Editing and Generation [Paper]
    • ORION: Decoupling and Alignment for Unified Autoregressive Understanding and Generation [Paper]
    • PCPO: Proportionate Credit Policy Optimization for Preference Alignment of Image Generation Models [Paper]
    • Reconstruction Alignment Improves Unified Multimodal Models [Paper]
    • Soft-Di[M]O: Improving One-Step Discrete Image Generation with Soft Embeddings [Paper]
    • SPEED: Scalable, Precise, and Efficient Concept Erasure for Diffusion Models [Paper]
    • STEER AWAY FROM MODE COLLISIONS: IMPROVING COMPOSITION IN DIFFUSION MODELS [Paper]
    • Story-Iter: A Training-free Iterative Paradigm for Long Story Visualization [Paper]
    • VLM-Guided Adaptive Negative Prompting for Creative Generation [Paper]
    • VSF: Simple, Efficient, and Effective Negative Guidance in Few-Step Image Generation Models By Value Sign Flip [Paper]
    • NextStep-1: Toward Autoregressive Image Generation with Continuous Tokens at Scale [Paper]
    • Rethinking Global Text Conditioning in Diffusion Transformers [Paper]
    • Forget Many, Forget Right: Scalable and Precise Concept Unlearning in Diffusion Models [Paper]
    • CASteer: Cross-Attention Steering for Controllable Concept Erasure [Paper]
    • CoEmoGen: Towards Semantically-Coherent and Scalable Emotional Image Content Generation [Paper]
    • Dynamic Classifier-Free Diffusion Guidance via Online Feedback [Paper]
    • Evolutionary Caching to Accelerate Your Off-the-Shelf Diffusion Model [Paper]
    • Relational Feature Caching for Accelerating Diffusion Transformers [Paper]
    • ScalingCache: Extreme Acceleration of DiTs through Difference Scaling and Dynamic Interval Caching [Paper]
    • RAPID$^3$: Tri-Level Reinforced Acceleration Policies for Diffusion Transformer [Paper]
    • Plug-and-Play Fidelity Optimization for Diffusion Transformer Acceleration via Cumulative Error Minimization [Paper]
    • DenseGRPO: From Sparse to Dense Reward for Flow Matching Model Alignment [Paper]
    • DiCache: Let Diffusion Model Determine Its Own Cache [Paper]
    • Towards Better Optimization For Listwise Preference in Diffusion Models [Paper]
    • FlowCast: Trajectory Forecasting for Scalable Zero-Cost Speculative Flow Matching [Paper]
    • HiCache: A Plug-in Scaled-Hermite Upgrade for Taylor-Style Cache-then-Forecast Diffusion Acceleration [Paper]
    • TwinFlow: Realizing One-step Generation on Large Models with Self-adversarial Flows [Paper]
    • OBS-Diff: Accurate Pruning For Diffusion Models in One-Shot [Paper]
    • Scale-wise Distillation of Diffusion Models [Paper]
    • DiffSparse: Accelerating Diffusion Transformers with Learned Token Sparsity [Paper]
    • ERTACache: Error Rectification and Timesteps Adjustment for Efficient Diffusion [Paper]
    • MeanCache: From Instantaneous to Average Velocity for Accelerating Flow Matching Inference [Paper]
    • Diffusion Negative Preference Optimization Made Simple [Paper]
    • Learnable Sparsity for Vision Generative Models [Paper]
    • Stochastic Self-Guidance for Training-Free Enhancement of Diffusion Models [Paper]
    • TEMPFLOW-GRPO: WHEN TIMING MATTERS FOR GRPO IN FLOW MODELS [Paper]
    • TreeGRPO: Tree-Advantage GRPO for Online RL Post-Training of Diffusion Models [Paper]
    • GLASS Flows: Efficient Inference for Reward Alignment of Flow and Diffusion Models [Paper]
    • Consistent Noisy Latent Rewards for Trajectory Preference Optimization in Diffusion Models [Paper]
    • BranchGRPO: Stable and Efficient GRPO with Structured Branching in Diffusion Models [Paper]
    • Scaling Group Inference for Diverse and High-Quality Generation [Paper]
    • FastFlow: Accelerating The Generative Flow Matching Models with Bandit Inference [Paper]
    • Generation then Reconstruction: Accelerating Masked Autoregressive Models via Two-Stage Sampling [Paper]
    • Constantly Improving Image Models Need Constantly Improving Benchmarks [Paper]
    • Closing the Safety Gap: Surgical Concept Erasure in Visual Autoregressive Models [Paper]
    • Exploring the Design Space of Transition Matching [Paper]
    • Synthetic History: Evaluating Visual Representations of the Past in Diffusion Models [Paper]
    • The Intricate Dance of Prompt Complexity, Quality, Diversity and Consistency in T2I Models [Paper]
    • Implicit Inversion turns CLIP into a Decoder [Paper]
    • DualToken: Towards Unifying Visual Understanding and Generation with Dual Visual Vocabularies [Paper]
    • Uni-X: Mitigating Modality Conflict with a Two-End-Separated Architecture for Unified Multimodal Models [Paper]
    • Uni-CoT: Towards Unified Chain-of-Thought Reasoning Across Text and Vision [Paper]
    • GoT-R1: Unleashing Reasoning Capability of Autoregressive Visual Generation with Reinforcement Learning [Paper]
    • Massive Activations are the Key to Local Detail Synthesis in Diffusion Transformers [Paper]
    • Let Features Decide Their Own Solvers: Hybrid Feature Caching for Diffusion Transformers [Paper]
    • K-Sort Eval: Efficient Preference Evaluation for Visual Generation via Corrected VLM-as-a-Judge [Paper]
  • AAAI
    • Plot’n Polish: Zero-Shot Story Visualization and Disentangled Editing with Text-to-Image Diffusion Models [Paper]
    • DOS: Directional Object Separation in Text Embeddings for Multi-Object Image Generation [Paper]
    • FastFLUX: Pruning FLUX with Block-wise Replacement and Sandwich Training [Paper]
    • Sortblock: Similarity-Aware Feature Reuse for Diffusion Model [Paper]
    • SAGA: Learning Signal-Aligned Distributions for Improved Text-to-Image Generation [Paper]
    • EvalMuse-40K: A Fine-Grained Benchmark with Comprehensive Human Annotations for Text-to-Image Generation Model Alignment Evaluation [Paper]
    • Margin-Aware Preference Optimization for Aligning Diffusion Models Without Reference [Paper]
    • Rethinking Direct Preference Optimization in Diffusion Models [Paper]
    • Towards Spatially Consistent Image Generation: On Incorporating Intrinsic Scene Properties into Diffusion Models [Paper]
    • MoSs: Mixture of Scales for Efficient High-Resolution Autoregressive Image Generation [Paper]
    • SpecDiff: Accelerating Diffusion Model Inference with Self-Speculation [Paper]
    • Infinite-Story: A Training-Free Consistent Text-to-Image Generation [Paper]
    • SphereDiff: Tuning-free 360° Static and Dynamic Panorama Generation via Spherical Latent Representation [Paper]
    • UniAlignment: Semantic Alignment for Unified Image Generation, Understanding, Manipulation and Perception [Paper]
    • Mass Concept Erasure in Diffusion Models with Concept Hierarchy [Paper]
    • Self-NPO: Data-Free Diffusion Model Enhancement via Truncated Diffusion Fine-Tuning [Paper]
    • Unleashing the Potential of Large Language Models for Text-to-Image Generation Through Autoregressive Representation Alignment [Paper]
    • VisionReward: Fine-Grained Multi-Dimensional Human Preference Learning for Image and Video Generation [Paper]
    • LongT2IBench: A Benchmark for Evaluating Long Text-to-Image Generation with Graph-structured Annotations [Paper]
    • RealUHR: Harnessing Patch-Cascade Flows for Photorealistic Ultra-High-Resolution Synthesis [Paper]
    • Forecast Then Calibrate: Feature Caching as ODE for Efficient Diffusion Transformers [Paper]
    • Exploring Position Encoding Mechanism in Diffusion U-Net for Training-free High-resolution Image Generation [Paper]
    • LookFlow: Training-Free and Efficient High-Resolution Image Synthesis via Dynamic Lookahead Guidance Flow [Paper]
    • Mechanistic Dissection of Cross-Attention Subspaces in Text-to-Image Diffusion Models [Paper]
    • ProCache: Constraint-Aware Feature Caching with Selective Computation for Diffusion Transformer Acceleration [Paper]
    • HierarchicalPrune: Position-Aware Compression for Large-Scale Diffusion Models [Paper]
    • Annealed Relaxation of Speculative Decoding for Faster Autoregressive Image Generation [Paper]
    • What Makes a Good Generated Image? Investigating Human and Multimodal LLM Image Preference Alignment [Paper]
    • Head-Aware KV Cache Compression for Efficient Visual Autoregressive Modeling [Paper]
    • FUSE: Fine-Grained and Semantic-Aware Learning for Unified Image Understanding and Generation [Paper]
    • Talk2Image: A Multi-Agent System for Multi-Turn Image Generation and Editing [Paper]
• Year 2025
  • CVPR
    • PreciseCam: Precise Camera Control for Text-to-Image Generation [Paper] [Project] [Code]
    • Type-R: Automatically Retouching Typos for Text-to-Image Generation [Paper] [Code]
    • Compass Control: Multi Object Orientation Control for Text-to-Image Generation [Paper]
    • Generative Photography: Scene-Consistent Camera Control for Realistic Text-to-Image Synthesis [Paper] [Project] [Code]
    • One-Way Ticket: Time-Independent Unified Encoder for Distilling Text-to-Image Diffusion Models [Paper] [Code]
    • Build a Text-to-Image Generator (from Scratch): Build your own vision transformer and diffusion models for text-to-image generation–from scratch! [Paid book]
    • Text Embedding is Not All You Need: Attention Control for Text-to-Image Semantic Alignment with Text Self-Attention Maps [Paper]
    • Towards Understanding and Quantifying Uncertainty for Text-to-Image Generation [Paper]
    • Plug-and-Play Interpretable Responsible Text-to-Image Generation via Dual-Space Multi-facet Concept Control [Paper] [Project] [Code]
    • Make It Count: Text-to-Image Generation with an Accurate Number of Objects [Paper] [Project] [Code]
    • MCCD: Multi-Agent Collaboration-based Compositional Diffusion for Complex Text-to-Image Generation [Paper]
    • Rethinking Training for De-biasing Text-to-Image Generation: Unlocking the Potential of Stable Diffusion [Paper]
    • ShapeWords: Guiding Text-to-Image Synthesis with 3D Shape-Aware Prompts [Paper] [Project]
    • SnapGen: Taming High-Resolution Text-to-Image Models for Mobile Devices with Efficient Architectures and Training [Paper]
    • Spatial Transport Optimization by Repositioning Attention Map for Training-Free Text-to-Image Synthesis [Paper]
    • Focus-N-Fix: Region-Aware Fine-Tuning for Text-to-Image Generation [Paper]
    • SILMM: Self-Improving Large Multimodal Models for Compositional Text-to-Image Generation [Paper]
    • Localized Concept Erasure for Text-to-Image Diffusion Models Using Training-Free Gated Low-Rank Adaptation [Paper]
    • Self-Cross Diffusion Guidance for Text-to-Image Synthesis of Similar Subjects [Paper]
    • Noise Diffusion for Enhancing Semantic Faithfulness in Text-to-Image Synthesis [Paper] [Code]
    • Learning to Sample Effective and Diverse Prompts for Text-to-Image Generation [Paper]
    • STEREO: A Two-Stage Framework for Adversarially Robust Concept Erasing from Text-to-Image Diffusion Models [Paper]
    • Minority-Focused Text-to-Image Generation via Prompt Optimization [Paper] [Code]
    • Scaling Down Text Encoders of Text-to-Image Diffusion Models [Paper] [Code]
    • Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation [Paper]
    • Enhancing Creative Generation on Stable Diffusion-based Models [Paper]
    • Aesthetic Post-Training Diffusion Models from Generic Preferences with Step-by-step Preference Optimization [Paper]
    • Stretching Each Dollar: Diffusion Training from Scratch on a Micro-Budget [Paper]
    • Diffusion-4K: Ultra-High-Resolution Image Synthesis with Latent Diffusion Models [Paper]
    • Precise, Fast, and Low-cost Concept Erasure in Value Space: Orthogonal Complement Matters [Paper]
    • T2ISafety: Benchmark for Assessing Fairness, Toxicity, and Privacy in Image Generation [Paper]
    • ChatGen: Automatic Text-to-Image Generation From FreeStyle Chatting [Paper]
    • Dual Diffusion for Unified Image Generation and Understanding [Paper]
    • DyMO: Training-Free Diffusion Model Alignment with Dynamic Multi-Objective Scheduling [Paper]
    • Erasing Undesirable Influence in Diffusion Models [Paper]
    • Redefining <Creative> in Dictionary: Towards an Enhanced Semantic Understanding of Creative Generation [Paper]
    • Six-CD: Benchmarking Concept Removals for Text-to-image Diffusion Models [Paper]
    • NitroFusion: High-Fidelity Single-Step Diffusion through Dynamic Adversarial Training [Paper]
    • ReNeg: Learning Negative Embedding with Reward Guidance [Paper]
    • UniReal: Universal Image Generation and Editing via Learning Real-world Dynamics [Paper]
    • OmniFlow: Any-to-Any Generation with Multi-Modal Rectified Flows [Paper]
    • FAM Diffusion: Frequency and Attention Modulation for High-Resolution Image Generation with Stable Diffusion [Paper]
    • DreamOmni: Unified Image Generation and Editing [Paper]
    • Curriculum Direct Preference Optimization for Diffusion and Consistency Models [Paper]
    • Dissecting and Mitigating Diffusion Bias via Mechanistic Interpretability [Paper]
    • ACE: Anti-Editing Concept Erasure in Text-to-Image Models [Paper]
    • One Diffusion to Generate Them All [Paper]
    • Let's Verify and Reinforce Image Generation Step by Step [Paper]
    • JanusFlow: Harmonizing Autoregression and Rectified Flow for Unified Multimodal Understanding and Generation [Paper]
    • Schedule On the Fly: Diffusion Time Prediction for Faster and Better Image Generation [Paper]
    • Fine-Grained Erasure in Text-to-Image Diffusion-based Foundation Models [Paper]
    • A Comprehensive Study of Decoder-Only LLMs for Text-to-Image Generation [Paper]
    • Exploring Sparse MoE in GANs for Text-conditioned Image Synthesis [Paper]
    • WeGen: A Unified Model for Interactive Multimodal Generation as We Chat [Paper]
    • VerbDiff: Text-Only Diffusion Models with Enhanced Interaction Awareness [Paper]
    • Memories of Forgotten Concepts [Paper]
    • PQPP: A Joint Benchmark for Text-to-Image Prompt and Query Performance Prediction [Paper]
    • Reward Fine-Tuning Two-Step Diffusion Models via Learning Differentiable Latent-Space Surrogate Reward [Paper]
    • Calibrated Multi-Preference Optimization for Aligning Diffusion Models [Paper]
    • InPO: Inversion Preference Optimization with Reparametrized DDIM for Efficient Diffusion Model Alignment [Paper]
    • Enhancing Privacy-Utility Trade-offs to Mitigate Memorization in Diffusion Models [Paper]
    • Spherical Manifold Guided Diffusion Model for Panoramic Image Generation [Paper]
    • Towards Transformer-Based Aligned Generation with Self-Coherence Guidance [Paper]
    • Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis [Paper]
    • Science-T2I: Addressing Scientific Illusions in Image Synthesis [Paper]
    • Towards Better Alignment: Training Diffusion Models with Reinforcement Learning Against Sparse Rewards [Paper]
    • ⚠️ The Illusion of Unlearning: The Unstable Nature of Machine Unlearning in Text-to-Image Diffusion Models [Paper]
    • ⚠️ Exploring the Deep Fusion of Large Language Models and Diffusion Transformers for Text-to-Image Synthesis [Paper]
    • ⚠️ Detect-and-Guide: Self-regulation of Diffusion Models for Safe Text-to-Image Generation via Guideline Token Optimization [Paper]
    • ⚠️ Multi-Group Proportional Representations for Text-to-Image Models [Paper]
    • ⚠️ VODiff: Controlling Object Visibility Order in Text-to-Image Generation [Paper]
  • ICCV
    • SANA-Sprint: One-Step Diffusion with Continuous-Time Consistency Distillation [Paper]
    • Erasing More Than Intended? How Concept Erasure Degrades the Generation of Non-Target Concepts [Paper]
    • Adversarial Distribution Matching for Diffusion Distillation Towards Efficient Image and Video Synthesis [Paper]
    • Golden Noise for Diffusion Models: A Learning Framework [Paper]
    • What Makes for Text to 360-degree Panorama Generation with Stable Diffusion? [Paper]
    • EDiT: Efficient Diffusion Transformers with Linear Compressed Attention [Paper]
    • TRCE: Towards Reliable Malicious Concept Erasure in Text-to-Image Diffusion Models [Paper]
    • Enhancing Reward Models for High-quality Image Generation: Beyond Text-Image Alignment [Paper]
    • Cycle Consistency as Reward: Learning Image-Text Alignment without Human Preferences [Paper]
    • CAP: Evaluation of Persuasive and Creative Image Generation [Paper]
    • FairGen: Enhancing Fairness in Text-to-Image Diffusion Models via Self-Discovering Latent Directions [Paper]
    • Trade-offs in Image Generation: How Do Different Dimensions Interact? [Paper]
    • TCFG: Truncated Classifier-Free Guidance for Efficient and Scalable Text-to-Image Acceleration [Paper]
    • Dense2MoE: Restructuring Diffusion Transformer to MoE for Efficient Text-to-Image Generation [Paper]
    • Holistic Unlearning Benchmark: A Multi-Faceted Evaluation for Text-to-Image Diffusion Model Unlearning [Paper]
    • Fair Generation without Unfair Distortions: Debiasing Text-to-Image Generation with Entanglement-Free Attention [Paper]
    • DIMCIM: A Quantitative Evaluation Framework for Default-mode Diversity and Generalization in Text-to-Image Generative Models [Paper]
    • Rethinking Cross-Modal Interaction in Multimodal Diffusion Transformers [Paper]
    • UniVG: A Generalist Diffusion Model for Unified Image Generation and Editing [Paper]
    • FlowTok: Flowing Seamlessly Across Text and Image Tokens [Paper]
    • From Reflection to Perfection: Scaling Inference-Time Optimization for Text-to-Image Diffusion Models via Reflection Tuning [Paper]
    • CuRe: Cultural Gaps in the Long Tail of Text-to-Image Systems [Paper]
    • Penalizing Boundary Activation for Object Completeness in Diffusion Models [Paper]
    • Decoding Correlation-Induced Misalignment in the Stable Diffusion Workflow for Text-to-Image Generation [Paper]
    • Adaptive Routing of Text-to-Image Generation Requests Between Large Cloud Model and Light-Weight Edge Model [Paper]
    • SuMa: A Subspace Mapping Approach for Robust and Effective Concept Erasure in Text-to-Image Diffusion Models [Paper]
    • Leveraging Panoptic Scene Graph for Evaluating Fine-Grained Text-to-Image Generation [Paper]
    • Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Concepts [Paper]
    • Sculpting Memory: Multi-Concept Forgetting in Diffusion Models via Dynamic Mask and Concept-Aware Optimization [Paper]
    • LLM Thought Divergence and Convergence for Dialogue-Based Image Generation Control [Paper]
    • Imbalance in Balance: Online Concept Balancing in Generation Models [Paper]
    • Multimodal LLMs as Customized Reward Models for Text-to-Image Generation [Paper]
    • The Silent Assistant: NoiseQuery as Implicit Guidance for Goal-Driven Image Generation [Paper]
    • ShortFT: Diffusion Model Alignment via Shortcut-based Fine-Tuning [Paper]
    • Rethinking DPO-style Diffusion Aligning Frameworks [Paper]
    • VSC: Visual Search Compositional Text-to-Image Diffusion Model [Paper]
    • Transformed Low-rank Adaptation via Tensor Decomposition and Its Applications to Text-to-image Models [Paper]
    • InfGen: A Resolution-Agnostic Paradigm for Scalable Image Synthesis [Paper]
    • FreeScale: Unleashing the Resolution of Diffusion Models via Tuning-Free Scale Fusion [Paper]
    • Text Embedding Knows How to Quantize Text-Guided Diffusion Models [Paper]
    • HPSv3: Towards Wide-Spectrum Human Preference Score [Paper]
    • DC-AR: Efficient Masked Autoregressive Image Generation with Deep Compression Hybrid Tokenizer [Paper]
    • DiTFastAttnV2: Head-wise Attention Compression for Multi-Modality Diffusion Transformers [Paper]
    • IQA-Adapter: Exploring Knowledge Transfer from Image Quality Assessment to Diffusion-based Generative Models [Paper]
    • Guiding Diffusion Models with Adaptive Negative Sampling Without External Resources [Paper]
    • AlignGuard: Scalable Safety Alignment for Text-to-Image Generation [Paper]
    • ICE-Bench: A Unified and Comprehensive Benchmark for Image Creating and Editing [Paper]
    • X2I: Seamless Integration of Multimodal Understanding into Diffusion Transformer via Attention Distillation [Paper]
    • T2I-Copilot: A Training-Free Multi-Agent Text-to-Image System for Enhanced Prompt Interpretation and Interactive Generation [Paper]
    • Lumina-Image 2.0: A Unified and Efficient Image Generative Framework [Paper]
    • Scendi Score: Prompt-Aware Diversity Evaluation via Schur Complement of CLIP Embeddings [Paper]
    • PUMA: Empowering Unified MLLM with Multi-granular Visual Generation [Paper]
    • Democratizing Text-to-Image Masked Generative Models with Compact Text-Aware One-Dimensional Tokens [Paper]
    • Reflect-DiT: Inference-Time Scaling for Text-to-Image Diffusion Transformers via In-Context Reflection [Paper]
    • Supercharged One-step Text-to-Image Diffusion Models with Negative Prompts [Paper]
    • IMG: Calibrating Diffusion Models via Implicit Multimodal Guidance [Paper]
    • FairHuman: Boosting Hand and Face Quality in Human Image Generation with Minimum Potential Delay Fairness in Diffusion Models [Paper]
    • Importance-Based Token Merging for Efficient Image and Video Generation [Paper]
    • Scale Your Instructions: Enhance the Instruction-Following Fidelity of Unified Image Generation Model by Self-Adaptive Attention Scaling [Paper]
    • Translation of Text Embedding via Delta Vector to Suppress Strongly Entangled Content in Text-to-Image Diffusion Models [Paper]
    • Dual-Process Image Generation [Paper]
    • LMM4LMM: Benchmarking and Evaluating Large-multimodal Image Generation with LMMs [Paper]
    • Reusing Computation in Text-to-Image Diffusion for Efficient Generation of Image Sets [Paper]
    • Scalable Ranked Preference Optimization for Text-to-Image Generation [Paper]
    • Frequency-Aware Autoregressive Modeling for Efficient High-Resolution Image Synthesis [Paper]
    • PLADIS: Pushing the Limits of Attention in Diffusion Models at Inference Time by Leveraging Sparsity [Paper]
  • NeurIPS
    • Rectified-CFG++ for Flow Based Models [Paper] [Code]
    • Mitigating Sexual Content Generation via Embedding Distortion in Text-conditioned Diffusion Models [Paper]
    • Adjusting Initial Noise to Mitigate Memorization in Text-to-Image Diffusion Models [Paper]
    • Self-Supervised Direct Preference Optimization for Text-to-Image Diffusion Models [Paper]
    • UltraHR-100K: Enhancing UHR Image Synthesis with A Large-Scale High-Quality Dataset [Paper] [Code]
    • Localizing Knowledge in Diffusion Transformers [Paper]
    • When Are Concepts Erased From Diffusion Models? [Paper]
    • Towards Resilient Safety-driven Unlearning for Diffusion Models against Downstream Fine-tuning [Paper] [Code]
    • Exploring Diffusion Transformer Designs via Grafting [Paper]
    • Moment- and Power-Spectrum-Based Gaussianity Regularization for Text-to-Image Models [Paper]
    • Scaling Diffusion Transformers Efficiently via $\mu$P [Paper]
    • LightFair: Towards an Efficient Alternative for Fair T2I Diffusion via Debiasing Pre-trained Text Encoders [Paper] [Code]
    • LEDiT: Your Length-Extrapolatable Diffusion Transformer without Positional Encoding [Paper] [Project]
    • Fourier Token Merging: Understanding and Capitalizing Frequency Domain for Efficient Image Generation [Paper] [Code]
    • Aligning Text-to-Image Diffusion Models to Human Preference by Classification [Paper]
    • AR-RAG: Autoregressive Retrieval Augmentation for Image Generation [Paper]
    • Unveiling Concept Attribution in Diffusion Models [Paper] [Code]
    • FairImagen: Post-Processing for Bias Mitigation in Text-to-Image Models [Paper] [Code]
    • Enhancing Text-to-Image Diffusion Transformer via Split-Text Conditioning [Paper]
    • Inference-Time Scaling for Flow Models via Stochastic Generation and Rollover Budget Forcing [Paper]
    • Towards Better & Faster Autoregressive Image Generation: From the Perspective of Entropy [Paper]
    • ObCLIP: Oblivious CLoud-Device Hybrid Image Generation with Privacy Preservation [Paper]
    • Ambient Diffusion Omni: Training Good Models with Bad Data [Paper]
    • T2I-R1: Reinforcing Image Generation with Collaborative Semantic-level and Token-level CoT [Paper]
    • Flow-GRPO: Training Flow Matching Models via Online RL [Paper]
    • RespoDiff: Dual-Module Bottleneck Transformation for Responsible & Faithful T2I Generation [Paper]
    • Entropy Rectifying Guidance for Diffusion and Flow Models [Paper]
    • Show-o2: Improved Native Unified Multimodal Models [Paper] [Code]
    • LABridge: Text–Image Latent Alignment Framework via Mean-Conditioned OU Process [Paper]
    • Re-ttention: Ultra Sparse Visual Generation via Attention Statistical Reshape [Paper]
    • Ranking-based Preference Optimization for Diffusion Models from Implicit User Feedback [Paper] [Code]
    • RepLDM: Reprogramming Pretrained Latent Diffusion Models for High-Quality, High-Efficiency, High-Resolution Image Generation [Paper] [Code]
    • EraseFlow: Learning Concept Erasure Policies via GFlowNet-Driven Alignment [Paper]
    • Speculative Jacobi-Denoising Decoding for Accelerating Autoregressive Text-to-image Generation [Paper]
    • CURE: Concept Unlearning via Orthogonal Representation Editing in Diffusion Models [[Paper]](https://proceedings.neurips.cc/paper_files/paper/2025/file/