mlsanityData to 14 Oct 2026

OmniPaint: Mastering Object-Oriented Editing via Disentangled Insertion-Removal Inpainting

arXiv 2503.08677 ↗First posted 11 Mar 2025cs.CVcited by 39

On the map: Diffusion models → 3D, point cloud and scene generation

Abstract (arXiv)

Diffusion-based generative models have revolutionized object-oriented image editing, yet their deployment in realistic object removal and insertion remains hampered by challenges such as the intricate interplay of physical effects and insufficient paired training data. In this work, we introduce OmniPaint, a unified framework that re-conceptualizes object removal and insertion as interdependent processes rather than isolated tasks. Leveraging a pre-trained diffusion prior along with a progressive training pipeline comprising initial paired sample optimization and subsequent large-scale unpaired refinement via CycleFlow, OmniPaint achieves precise foreground elimination and seamless object insertion while faithfully preserving scene geometry and intrinsic properties. Furthermore, our novel CFD metric offers a robust, reference-free evaluation of context consistency and object hallucination, establishing a new benchmark for high-fidelity image editing. Project page: https://yeates.github.io/OmniPaint-Page/

Key facts. Cited by 39 papers in the map (data through 14 Oct 2026). Placed in 3D, point cloud and scene generation, one of 14 directions in Diffusion models. Builds on 53 of its 69 references that are in the map. First posted on arXiv 11 Mar 2025.

Builds on

53 of 69 references.

Closest ancestors · 6

Same direction, newest first

  1. 2024Baking Gaussian Splatting into Diffusion Denoiser for Fast and Scalable Single-stage Image-to-3D Generation and Reconstruction21 Nov
  2. 2024MagicEraser: Erasing Any Objects via Semantics-Aware Control14 Oct
  3. 2024ReMOVE: A Reference-free Metric for Object Erasure1 Sep
  4. 2024ObjectDrop: Bootstrapping Counterfactuals for Photorealistic Object Removal and Insertion27 Mar
  5. 2023Magicremover: Tuning-free Text-guided Image inpainting with Diffusion Models4 Oct
  6. 2023Inst-Inpaint: Instructing to Remove Objects with Diffusion Models6 Apr

Other references · 47

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  4. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  5. 2023Segment Anythingcited by 8,234
  6. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  7. 2022Flow Matching for Generative Modelingcited by 5,702
  8. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  9. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  10. 2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 3,700
  11. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Modelscited by 3,614
  12. 2021MUSIQ: Multi-scale Image Quality Transformercited by 1,081
35 more
  1. 2016COCO-Stuff: Thing and Stuff Classes in Contextcited by 864
  2. 2021Resolution-robust Large Mask Inpainting with Fourier Convolutionscited by 738
  3. 2022MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessmentcited by 522
  4. 2022Paint by Example: Exemplar-based Image Editing with Diffusion Modelscited by 489
  5. 2022Blended Latent Diffusioncited by 432
  6. 2023AnyDoor: Zero-shot Object-level Image Customizationcited by 412
  7. 2023Segment Anything in High Qualitycited by 376
  8. 2024OminiControl: Minimal and Universal Control for Diffusion Transformercited by 340
  9. 2023Rethinking FID: Towards a Better Evaluation Metric for Image Generationcited by 310
  10. 2022MAT: Mask-Aware Transformer for Large Hole Image Inpaintingcited by 218
  11. 2023Guiding Instruction-based Image Editing via Multimodal Large Language Modelscited by 185
  12. 2023A Task is Worth One Word: Learning with Task Prompts for High-Quality Versatile Image Inpaintingcited by 169
  13. 2017Deep Image Harmonizationcited by 120
  14. 2025VideoPainter: Any-length Video Inpainting and Editing with Plug-and-Play Context Controlcited by 84
  15. 2017GP-GAN: Towards Realistic High-Resolution Image Blendingcited by 64
  16. 2024IMPRINT: Generative Object Compositing by Learning Identity-Preserving Representationcited by 64
  17. 2022CM-GAN: Image Inpainting with Cascaded Modulation GAN and Object-Aware Trainingcited by 49
  18. 2024Large-Scale Text-to-Image Model with Inpainting is a Zero-Shot Subject-Driven Image Generatorcited by 48
  19. 2024Add-it: Training-Free Object Insertion in Images With Pretrained Diffusion Modelscited by 43
  20. 2024PromptFix: You Prompt and We Fix the Photocited by 40
  21. 2024MULAN: A Multi Layer Annotated Dataset for Controllable Text-to-Image Generationcited by 38
  22. 2024TANGO: Co-Speech Gesture Video Reenactment with Hierarchical Audio Motion Embedding and Diffusion Interpolationcited by 32
  23. 2024BrushEdit: All-In-One Image Inpainting and Editingcited by 31
  24. 2024Open-Vocabulary Attention Maps with Token Optimization for Semantic Segmentation in Diffusion Modelscited by 29
  25. 2025SmartEraser: Remove Anything from Images using Masked-Region Guidancecited by 24
  26. 2025RORem: Training a Robust Object Remover with Human-in-the-Loopcited by 23
  27. 2023Diverse Inpainting and Editing with GAN Inversioncited by 20
  28. 2023Are These the Same Apple? Comparing Images Based on Object Intrinsicscited by 19
  29. 2024FreeCompose: Generic Zero-Shot Image Composition with Diffusion Priorcited by 16
  30. 2024ObjectMate: A Recurrence Prior for Object Insertion and Subject-Driven Generationcited by 16
  31. 2022High-Fidelity Image Inpainting with GAN Inversioncited by 15
  32. 2023MaGIC: Multi-modality Guided Image Completioncited by 14
  33. 2024Exploring Phrase-Level Grounding with Text-to-Image Diffusion Modelcited by 13
  34. 2022Unbiased Multi-Modality Guidance for Image Inpaintingcited by 10
  35. 2024MetaShadow: Object-Centered Shadow Detection, Removal, and Synthesiscited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 15; 2026: 24.

39 papers.

Sort:

By area:

  1. 2025NoHumansRequired: Autonomous High-Quality Image Editing Triplet Miningcited by 33
  2. 2025Does FLUX Already Know How to Perform Physically Plausible Image Composition?cited by 32
  3. 2025PhyGDPO: Physics-Aware Groupwise Direct Preference Optimization for Physically Consistent Text-to-Video Generationcited by 23
  4. 2025MMIG-Bench: Towards Comprehensive and Explainable Evaluation of Multi-Modal Image Generation Modelscited by 17
  5. 2025OmniVCus: Feedforward Subject-driven Video Customization with Multimodal Control Conditionscited by 17
  6. 2025MIRA: Multimodal Iterative Reasoning Agent for Image Editingcited by 17
  7. 2026EffectErase: Joint Video Object Removal and Insertion for High-Quality Effect Erasingcited by 15
  8. 2025Precise Object and Effect Removal with Adaptive Target-Aware Attentioncited by 13
  9. 2026MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agentscited by 11
  10. 2025GeoRemover: Removing Objects and Their Causal Visual Artifactscited by 9
  11. 2025In-Context Brush: Zero-shot Customized Subject Insertion with Context-Aware Latent Space Manipulationcited by 7
  12. 2026Credit Where It is Due: Cross-Modality Connectivity Drives Precise Reinforcement Learning for MLLM Reasoningcited by 6
27 more
  1. 2026From Ideal to Real: Stable Video Object Removal under Imperfect Conditionscited by 5
  2. 2026VideoArgus: Agentic Rubric-Grounded Unified Evaluation for Video Generation and Editingcited by 5
  3. 2025RePainter: Empowering E-commerce Object Removal via Spatial-matting Reinforcement Learningcited by 3
  4. 2026RoboDream: Compositional World Models for Scalable Robot Data Synthesiscited by 3
  5. 2025PixPerfect: Seamless Latent Diffusion Local Editing with Discriminative Pixel-Space Refinementcited by 2
  6. 2026VIBE: Visual Instruction Based Editorcited by 2
  7. 2025EasyOmnimatte: Taming Pretrained Inpainting Diffusion Models for End-to-End Video Layered Decompositioncited by 1
  8. 2026FlashClear: Ultra-Fast Image Content Removal via Efficient Step Distillation and Feature Cachingcited by 1
  9. 2026DORS: Dynamic Attention Routing for Diffusion-based Object Removal in Dense Scenescited by 1
  10. 2026BeyondMasks: Evaluating Causal and Physical Consistency in Video Object Removalcited by 1
  11. 2026EpiCon: Collective Agent Learning through Co-Evolving Multimodal Memorycited by 1
  12. 2025Mask Consistency Regularization in Object Removalcited by 0
  13. 2025InsertAnywhere: Geometrically Grounded and Optics-Aware Video Object Insertioncited by 0
  14. 2025EraseLoRA: MLLM-Driven Foreground Exclusion and Background Subtype Aggregation for Dataset-Free Object Removalcited by 0
  15. 2026Aligned Stable Inpainting: Mitigating Unwanted Object Insertion and Preserving Color Consistencycited by 0
  16. 2026Consistency-Preserving Concept Erasure via Unsafe-Safe Pairing and Directional Fisher-weighted Adaptationcited by 0
  17. 2026PICS: Pairwise Image Compositing with Spatial Interactionscited by 0
  18. 2026You Only Erase Once: Erasing Anything without Bringing Unexpected Contentcited by 0
  19. 2026Reflection Generation for Composite Image Using Diffusion Modelcited by 0
  20. 2026CatalogStitch: Dimension-Aware and Occlusion-Preserving Object Compositing for Catalog Image Generationcited by 0
  21. 2026AdaEraser: Training-Free Object Removal via Adaptive Attention Suppressioncited by 0
  22. 2026RefGC-SR2: Reference-guided Super-Resolution and Refinement of AI Generated Contentcited by 0
  23. 2026RoPEMover: Depth-Aware Object Relocation via Positional Embeddingscited by 0
  24. 2026OSOR: One-Step Diffusion Inpainting for Effect-Aware Object Removalcited by 0
  25. 2026TurboClear: One-Step Object-Effect Removal via Region-Calibrated Distribution Matching and Fusioncited by 0
  26. 2026Mover360: Controllable Object Manipulation in 360deg Panoramic Imagescited by 0
  27. 2026StreamErase: Towards Ultra-Fast Streaming Video Erasingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Paint by Inpaint: Learning to Add Image Objects by Removing Them First
  2. 2025OmnimatteZero: Fast Training-free Omnimatte with Pre-trained Video Diffusion Models
  3. 2025CrimEdit: Controllable Editing for Counterfactual Object Removal, Insertion, and Movement
  4. 2024Diffree: Text-Guided Shape Free Object Inpainting with Diffusion Model
  5. 2026Editing on the Generative Manifold: A Theoretical and Empirical Study of General Diffusion-Based Image Editing Trade-offs
  6. 2022Imagen Editor and EditBench: Advancing and Evaluating Text-Guided Image Inpainting
  7. 2024Generative Omnimatte: Learning to Decompose Video into Layers
  8. 2025GuidPaint: Class-Guided Image Inpainting with Diffusion Models