mlsanityData to 7 Oct 2026

Making Images Real Again: A Comprehensive Survey on Deep Image Composition

arXiv 2106.14490 ↗First posted 28 Jun 2021cs.CVcited by 63

Not on the map. Why?

Abstract (arXiv)

As a common image editing operation, image composition/compositing, which is also called object/subject insertion/addition/compositing, aims to combine the foreground from one image and another background image to produce a composite image. However, there are many issues that could make the composite images unrealistic. These issues can be summarized as the inconsistency between foreground and background, which includes appearance inconsistency, geometry inconsistency, and semantic inconsistency. The image composition task could be decomposed into multiple sub-tasks, in which each sub-task targets one or more issues. Specifically, object placement aims to find reasonable scale, location, and shape for the foreground. Image blending aims to address the unnatural boundary between foreground and background. Image harmonization aims to adjust the illumination statistics of foreground. Shadow (resp., reflection) generation aims to generate plausible shadow (resp., reflection) for the foreground. These sub-tasks can be executed sequentially or in parallel to acquire realistic composite images. To the best of our knowledge, there is no previous survey on image composition. In this paper, we conduct a comprehensive survey over the sub-tasks and combined task of image composition. For each one, we summarize the existing methods, available datasets, and common evaluation metrics. Datasets and codes for image composition are summarized at https://github.com/bcmi/Awesome-Object-Insertion. We have also contributed the first image composition toolbox: libcom https://github.com/bcmi/libcom, which assembles 10+ image-composition-related functions. The ultimate goal of this toolbox is to solve all image composition problems with simple `import libcom'. Based on libcom toolbox, we also develop an online image composition workbench https://libcom.ustcnewly.com.

Key facts. Cited by 63 papers in the map (data through 7 Oct 2026). Builds on 107 of its 329 references that are in the map. First posted on arXiv 28 Jun 2021.

Builds on

107 of 329 references.

References · 107

  1. 2017Attention Is All You Needcited by 59,932
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  4. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  5. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,893
  6. 2016Image-to-Image Translation with Conditional Adversarial Networkscited by 4,510
  7. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  8. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  9. 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,580
  10. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,880
  11. 2018Large Scale GAN Training for High Fidelity Natural Image Synthesiscited by 2,609
  12. 2017Non-local Neural Networkscited by 2,290
38 more
  1. 2014Conditional Generative Adversarial Netscited by 2,152
  2. 2017Arbitrary Style Transfer in Real-time with Adaptive Instance Normalizationcited by 1,998
  3. 2015Spatial Transformer Networkscited by 1,423
  4. 2020Image Segmentation Using Deep Learning: A Surveycited by 659
  5. 2020Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentationcited by 595
  6. 2018Free-Form Image Inpainting with Gated Convolutioncited by 594
  7. 2018The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scalecited by 542
  8. 2018Image Inpainting for Irregular Holes Using Partial Convolutionscited by 505
  9. 2015A Neural Algorithm of Artistic Stylecited by 480
  10. 2016Semantic Scene Completion from a Single Depth Imagecited by 468
  11. 2017NIMA: Neural Image Assessmentcited by 387
  12. 2019Deep Learning for Image Super-resolution: A Surveycited by 368
  13. 2017Cut, Paste and Learn: Surprisingly Easy Synthesis for Instance Detectioncited by 209
  14. 2018Arbitrary Style Transfer with Style-Attentional Networkscited by 197
  15. 2016Deeply supervised salient object detection with short connectionscited by 196
  16. 2017Deep Photo Style Transfercited by 168
  17. 2017Learning to Predict Indoor Illumination from a Single Imagecited by 160
  18. 2018Deep Learning for Single Image Super-Resolution: A Brief Reviewcited by 158
  19. 2016Semantic Image Inpainting with Deep Generative Modelscited by 156
  20. 2018Learning Rich Features for Image Manipulation Detectioncited by 154
  21. 2017Deep Image Mattingcited by 150
  22. 2018Fighting Fake News: Image Splice Detection via Learned Self-Consistencycited by 136
  23. 2019DoveNet: Deep Image Harmonization via Domain Verificationcited by 134
  24. 2018A Closed-form Solution to Photorealistic Image Stylizationcited by 133
  25. 2019Deep Learning on Image Denoising: An overviewcited by 128
  26. 2016Photo Aesthetics Ranking Network with Attributes and Content Adaptationcited by 122
  27. 2017Deep Image Harmonizationcited by 120
  28. 2018Modeling Visual Context is Key to Augmenting Object Detection Datasetscited by 106
  29. 2020Self-Supervised Scene De-occlusioncited by 96
  30. 2019InstaBoost: Boosting Instance Segmentation via Probability Map Guided Copy-Pastingcited by 95
  31. 2020SPAN: Spatial Pyramid Attention Network forImage Manipulation Localizationcited by 92
  32. 2019Mask-ShadowGAN: Learning to Remove Shadows from Unpaired Datacited by 91
  33. 2018ST-GAN: Spatial Transformer Generative Adversarial Networks for Image Compositingcited by 90
  34. 2019Learning Physics-guided Face Relighting under Directional Lightcited by 90
  35. 2020Bridging Composite and Real: Towards End-to-end Deep Image Mattingcited by 89
  36. 2019Deep Parametric Indoor Lighting Estimationcited by 82
  37. 2019Improving the Harmony of the Composite Image by Spatial-Separated Attention Modulecited by 73
  38. 2019Fast Spatially-Varying Indoor Lighting Estimationcited by 72

50 most cited of 107 All 107 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 3; 2022: 7; 2023: 18; 2024: 18; 2025: 10; 2026: 7.

63 papers.

Sort:

By area:

  1. 2022Paint by Example: Exemplar-based Image Editing with Diffusion Modelscited by 489
  2. 2023TF-ICON: Diffusion-Based Training-Free Cross-Domain Image Compositioncited by 182
  3. 2024Neural Gaffer: Relighting Any Object via Diffusioncited by 91
  4. 2021High-Resolution Image Harmonization via Collaborative Dual Transformationscited by 55
  5. 2023ControlCom: Controllable Image Composition using Diffusion Modelcited by 43
  6. 2023DreamEdit: Subject-driven Image Editingcited by 37
  7. 20243DRealCar: An In-the-wild RGB-D Car Dataset with 360-degree Viewscited by 37
  8. 2023Neural Preset for Color Style Transfercited by 36
  9. 2025Does FLUX Already Know How to Perform Physically Plausible Image Composition?cited by 30
  10. 2021OPA: Object Placement Assessment Datasetcited by 28
  11. 2023Semi-supervised Parametric Real-world Image Harmonizationcited by 26
  12. 2022Learning Object Placement via Dual-path Graph Completioncited by 20
38 more
  1. 2024PrimeComposer: Faster Progressively Combined Diffusion for Image Composition with Attention Steeringcited by 20
  2. 2023SmartMask: Context Aware High-Fidelity Mask Generation for Fine-grained Object Insertion and Layout Controlcited by 18
  3. 2023DreamCom: Finetuning Text-guided Inpainting Model for Image Compositioncited by 17
  4. 2024Shadow Generation for Composite Image Using Diffusion modelcited by 17
  5. 2023TopNet: Transformer-based Object Placement Network for Image Compositingcited by 16
  6. 2023MMNet: Multi-Collaboration and Multi-Supervision Network for Sequential Deepfake Detectioncited by 15
  7. 2022Fast Object Placement Assessmentcited by 12
  8. 2023Deep Image Matting: A Comprehensive Surveycited by 12
  9. 2023Painterly Image Harmonization using Diffusion Modelcited by 12
  10. 2023Deep Image Harmonization in Dual Color Spacescited by 12
  11. 2023Scene-aware Human Pose Generation using Transformercited by 9
  12. 2024TALE: Training-free Cross-domain Image Composition via Adaptive Latent Manipulation and Energy-guided Optimizationcited by 9
  13. 2024Generative Artificial Intelligence Meets Synthetic Aperture Radar: A Surveycited by 9
  14. 2024Compose and Conquer: Diffusion-Based 3D Depth Aware Composable Image Synthesiscited by 7
  15. 2023Foreground Object Search by Distilling Composite Image Featurecited by 6
  16. 2024Repositioning the Subject within Imagecited by 6
  17. 2021SAC-GAN: Structure-Aware Image Compositioncited by 5
  18. 2023Deep Image Harmonization with Learnable Augmentationcited by 5
  19. 2024MetaShadow: Object-Centered Shadow Detection, Removal, and Synthesiscited by 5
  20. 2022GALA: Toward Geometry-and-Lighting-Aware Object Search for Compositingcited by 4
  21. 2022Scrape, Cut, Paste and Learn: Automated Dataset Generation Applied to Parcel Logisticscited by 4
  22. 2023Deep Image Harmonization with Globally Guided Feature Transformation and Relation Distillationcited by 4
  23. 2024Training-Free Semantic Video Composition via Pre-trained Diffusion Modelcited by 4
  24. 2024Boosting General Trimap-free Matting in the Real-World Imagecited by 4
  25. 2024Foodfusion: A Novel Approach for Food Image Composition via Diffusion Modelscited by 4
  26. 2025AIComposer: Any Style and Content Image Composition via Feature Integrationcited by 4
  27. 2025MagicQuillV2: Precise and Interactive Image Editing with Layered Visual Cuescited by 4
  28. 2024DiffPop: Plausibility-Guided Object Placement Diffusion for Image Compositioncited by 3
  29. 2025Object Placement for Anythingcited by 3
  30. 2022Spatial Transformation for Image Composition via Correspondence Learningcited by 2
  31. 2024WildFit: Autonomous In-situ Model Adaptation for Resource-Constrained IoT Systemscited by 2
  32. 2024Diverse Image Harmonizationcited by 1
  33. 2024Revisit Self-supervised Depth Estimation with Local Structure-from-Motioncited by 1
  34. 2024Local Lesion Generation is Effective for Capsule Endoscopy Image Data Augmentation in a Limited Data Settingcited by 1
  35. 2022CcHarmony: Color-checker based Image Harmonization Datasetcited by 0
  36. 2023EqGAN: Feature Equalization Fusion for Few-shot Image Generationcited by 0
  37. 2023Painterly Image Harmonization via Adversarial Residual Learningcited by 0
  38. 2024DepGAN: Leveraging Depth Maps for Handling Occlusions and Transparency in Image Compositioncited by 0

50 most cited of 63 All 63 (JSON)