Compositional Image Generation

Ask about this topic

Overview

The field aims to make image generators reliably follow instructions about what belongs in a scene, where it goes, how objects relate, and which details come from reference images. Approaches include layout- and graph-conditioned diffusion, spatial and occlusion controls, modular adapters and layered synthesis, and agentic planning with iterative generation and editing. Evaluation spans layout-to-image, multi-subject and multi-reference tasks, with benchmarks such as CogCanvas and TRACE-Bench probing identity, attribute binding, relations, and spatial accuracy. A central challenge is generalizing to unfamiliar combinations and crowded scenes: binding and identity errors grow with complexity, while stronger control must preserve image quality and remain efficient.

Latest papers

CardsList

85 papers11 papers in the last four weeks, up 120% on the four weeks before. 0.1% of all new papers.

  1. Multimodal Large Language Models for Multi-Subject In-Context Image Generation

    Apr 8, 2026Yucheng Zhou, Dubing Chen, Huan Zheng +1Multi-Subject Image GenerationImage Generation

    SKL-IOTSC, CIS, University of Macau

  2. Graph-PiT: Enhancing Structural Coherence in Part-Based Image Synthesis via Graph Priors

    Apr 7, 2026Junbin Zhang, Meng Cao, Feng Tan +2Geometry-Conditioned Image GenerationImage Generation

    Guangdong Provincial Key Laboratory of Ultra High Definition Immersive Media Technology Shenzhen Graduate School, Peking University, China · Department of Computer Vision, Mohamed bin Zayed University of Artificial Intelligence, Abu Dhabi, United Arab Emirates

  3. Controllable Image Generation with Composed Parallel Token Prediction

    Apr 7, 2026Jamie Stirling, Noura Al-Moubayed, Chris G. Willcocks +1Controllable Image GenerationCompositional Image Generation

    Durham University, United Kingdom

  4. PortraitCraft: A Benchmark for Portrait Composition Understanding and Generation

    Apr 4, 2026Yuyang Sha, Zijie Lou, Youyun Tang +6Compositional Image GenerationCompositional T2I Generation

    MT Lab, Meitu Inc · Airbrush Studio, Meitu Inc · Meitu Design, Meitu Inc

  5. Reflection Generation for Composite Image Using Diffusion Model

    Apr 2, 2026Haonan Zhao, Qingyang Liu, Jiaxuan Chen +1Compositional Image GenerationDiffusion-Based Image Generation

    Shanghai Jiao Tong University

  6. ConceptWeaver: Weaving Disentangled Concepts with Flow

    Mar 30, 2026Jintao Chen, Aiming Hao, Xiaoqing Chen +6Subject-Driven Image GenerationCompositional Image Generation

    Peking University · AMAP, Alibaba Group

  7. MACRO: Advancing Multi-Reference Image Generation with Structured Long-Context Data

    Mar 26, 2026Zhekai Chen, Yuqing Wang, Manyuan Zhang +1Image GenerationCompositional Image Generation

    HKU MMLab · Meituan

  8. EchoGen: Cycle-Consistent Learning for Unified Layout-Image Generation and Understanding

    Mar 18, 2026Kai Zou, Hongbo Liu, Dian Zheng +3Cycle-Consistent LearningImage Generation

    School of Cyber Science and Technology, University of Science and Technology of China · Anhui Province Key Laboratory of Digital Security · Tongji University +3

  9. LaDe: Unified Multi-Layered Graphic Media Generation and Decomposition

    Mar 18, 2026Vlad-Constantin Lungu-Stan, Ionut Mironica, Mariana-Iuliana GeorgescuImage GenerationLatent Diffusion Models

    Adobe Research, Romania

  10. CompDiff enables fair and zero shot medical image generation across demographic intersections through compositional diffusion

    Mar 17, 2026Mahmoud K. Ibrahim, Bart Elen, Chang Sun +5HealthcareCompositional Image Generation

    Institute of Data Science, Faculty of Science and Engineering, Maastricht University, Maastricht, The Netherlands · Department of Advanced Computing Sciences, Faculty of Science and Engineering, Maastricht University, Maastricht, The Netherlands · VITO, Belgium

  11. LICA: Layered Image Composition Annotations for Graphic Design Research

    Mar 17, 2026Elad Hirsch, Shubham Yadav, Mohit Garg +1Compositional Image Generation

  12. Multi-Object Advertisement Creative Generation

    Mar 14, 2026Jialu Gao, Mithun Das Gupta, Qun Li +4Image GenerationE-Commerce

    CMU, USA · Microsoft AI, USA · Microsoft Research, USA

  13. EmoStory: Emotion-Aware Story Generation

    Mar 11, 2026Jingyuan Yang, Rucong Chen, Weibin Luo +1Compositional Image Generation

    CSSE, Shenzhen University

  14. VSDiffusion: Taming Ill-Posed Shadow Generation via Visibility-Constrained Diffusion

    Mar 9, 2026Jing Li, Jing ZhangCompositional Image GenerationDiffusion-Based Image Editing

    East China University of Science and Technology, Shanghai, China

  15. PICS: Pairwise Image Compositing with Spatial Interactions

    Mar 6, 2026Hang Zhou, Xinxin Zuo, Sen Wang +1Image EditingCompositional Image Generation

    University of Alberta · Concordia University

  16. InnoAds-Composer: Efficient Condition Composition for E-Commerce Poster Generation

    Mar 6, 2026Yuxin Qin, Ke Cao, Haowei Liu +13Image GenerationCompositional Image Generation

    JD.com, Inc., Beijing, China · Chongqing University of Posts and Telecommunications, Chongqing, China · The Hong Kong University of Science and Technology, Hong Kong, China +1

  17. Order Is Not Layout: Order-to-Space Bias in Image Generation

    Mar 4, 2026Yongkang Zhang, Zonglin Zhao, Yuechen Zhang +3T2I GenerationCompositional Image Generation

    Renmin University of China, China · Huazhong Agricultural University, China · Huazhong University of Science and Technology, China +2

  18. Training-Free Multi-Concept Image Editing

    Feb 24, 2026Niki Foteinopoulou, Ignas Budvytis, Stephan LiwickiTraining-Free Image EditingImage Editing

    Cambridge Research Laboratory, Toshiba Europe, Cambridge, UK · Independent Researcher, UK

  19. RL-RIG: A Generative Spatial Reasoner via Intrinsic Reflection

    Feb 23, 2026Tianyu Wang, Zhiyuan Ma, Qian Wang +3Image GenerationControllable Image Generation

    Tianyu Wang is with the Zhiyuan College, Shanghai Jiao Tong University · Zhiyuan Ma is with the Huazhong University of Science and Technology · Qian Wang is with the National University of Singapore +3

  20. OSInsert: Towards High-authenticity and High-fidelity Image Composition

    Feb 23, 2026Jingyuan Wang, Li NiuCompositional Image Generation

    MoE Key Lab of Artificial Intelligence, Shanghai Jiao Tong University

  21. DEIG: Detail-Enhanced Instance Generation with Fine-Grained Semantic Control

    Feb 20, 2026Shiyan Du, Conghan Yue, Xinyu Cheng +1Image GenerationCompositional Image Generation

    Sun Yat-sen University, Guangzhou, China · Fudan University, Shanghai, China · Yale University, New Haven, United States

  22. REPA-G: Test-Time Conditioning with Representation-Aligned Visual Features

    Feb 3, 2026Nicolas Sereyjol-Garros, Ellington Kirby, Victor Letzelter +2Diffusion Model GuidanceConditional Image Generation

    Valeo.ai, Paris, France · LTCI, Télécom Paris, Institut Polytechnique de Paris, France

  23. CanvasComposer: Personalized Group Photo Generation via a Multi-Reference Canvas

    Oct 23, 2025Gordon Guocheng Qian, Ruihang Zhang, Tsai-Shien Chen +11Subject-Driven Image GenerationImage Generation

    Snap Inc., USA · University of Toronto, Canada · University of California, Merced, USA +1

  24. What Drives Compositional Generalization in Visual Generative Models? The Importance of Continuous Training Objectives

    Oct 3, 2025Karim Farid, Rajat Sahay, Yumna Ali Alnaggar +4Compositional Image GenerationGenerative Modeling

    University of Freiburg · Bosch Center for Artificial Intelligence · Inria, École Normale Supérieure, CNRS, PSL Research University

  25. CompArt: Operationalizing Aesthetic Alignment in Text-to-Image Generation via Principles of Art

    Mar 15, 2025Zhe Jin, Tat-Seng ChuaImage GenerationCompositional Image Generation

    National University of Singapore Singapore