图像编辑论文趋势 2023–2026

主要结论

• 图像编辑论文从 2023 年的 112 篇增长到 2026 年的 238 篇，三年间扩大到约 2.1 倍。
• CVPR 是发图像编辑论文最多的会议，2023–2026 年共 364 篇，占四个会议总量的 53%。
• 交叉方向里涨得最快的是应用 · computer vision：同时属于两个主题的论文从 2023 年的 1 篇增至 2026 年的 15 篇。
• 在「多模态生成」的 5 个主题里，它的 2026 年论文数排第 3；同组增长最快的叙事生成是 15 倍，图像编辑是 2.1 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026Accelerated Image Editing via Consistency-Aware Source Token Pruning
• ICML 2026Klein Hyperbolic Metric Learning
• ICLR 2026DiffVax: Optimization-Free Image Immunization Against Diffusion-Based Editing
• CVPR 2026AnchorFlow Training-Free 3D Editing via Latent Anchor-Aligned Flows
• NeurIPS 2025A$^3$E: Towards Compositional Model Editing
• ICML 2025DEALing with Image Reconstruction: Deep Attentive Least Squares
• ICLR 2025HQ-Edit: A High-Quality Dataset for Instruction-based Image Editing
• CVPR 2025AnyEdit: Mastering Unified High-Quality Image Editing for Any Idea
• NeurIPS 2024Reproducibility study of “LICO: Explainable Models with Language-Image Consistency"
• ICML 2024Counterfactual Image Editing
• ICLR 2024Guiding Instruction-based Image Editing via Multimodal Large Language Models
• CVPR 2024DiffEditor: Boosting Accuracy and Flexibility on Diffusion-based Image Editing
• NeurIPS 2023ImageBrush: Learning Visual In-Context Instructions for Exemplar-Based Image Manipulation
• ICML 2023Random Shuffle Transformer for Image Restoration
• ICLR 2023Editing models with task arithmetic
• CVPR 20233D GAN Inversion With Facial Symmetry Prior
• NeurIPS 2026AnyBokeh: Physics-Guided Any-to-Any Bokeh Editing with Optical Fingerprint Transfer
• ICML 2026DLEBench: Evaluating Small-scale Object Editing Ability for Instruction-based Image Editing Model
• ICLR 2026Energy-Regularized Sequential Model Editing on Hyperspheres
• CVPR 2026ChordEdit One-Step Low-Energy Transport for Image Editing

全部图像编辑论文

共 108 篇，第 1 / 3 页

1. ICLR 2026Charts Are Not Images: On the Challenges of Scientific Chart Editing
2. ICLR 2026Consistent Text-to-Image Generation via Scene De-Contextualization
3. ICLR 2026Dataless Weight Disentanglement in Task Arithmetic via Kronecker-Factored Approximate Curvature
4. ICLR 2026Deconstructing Guidance: A Semantic Hierarchy for Precise Diffusion Model Editing
5. ICLR 2026DiffVax: Optimization-Free Image Immunization Against Diffusion-Based Editing
6. ICLR 2026Distributional Consistency Loss: Beyond Pointwise Data Terms in Inverse Problems
7. ICLR 2026Does FLUX Already Know How to Perform Physically Plausible Image Composition?
8. ICLR 2026DragFlow: Unleashing DiT Priors with Region-Based Supervision for Drag Editing
9. ICLR 2026Dragging with Geometry: From Pixels to Geometry-Guided Image Editing
10. ICLR 2026EditReward: A Human-Aligned Reward Model for Instruction-Guided Image Editing
11. ICLR 2026EditScore: Unlocking Online RL for Image Editing via High-Fidelity Reward Modeling
12. ICLR 2026EditVerse: Unifying Image and Video Editing and Generation with In-Context Learning
13. ICLR 2026EdiVal-Agent: An Object-Centric Framework for Automated, Fine-Grained Evaluation of Multi-Turn Editing
14. ICLR 2026Energy-Regularized Sequential Model Editing on Hyperspheres
15. ICLR 2026Factuality Matters: When Image Generation and Editing Meet Structured Visuals
16. ICLR 2026FaSTA*: Fast-Slow Toolpath Agent with Subroutine Mining for Efficient Multi-turn Image Editing
17. ICLR 2026FlowAlign: Trajectory-Regularized, Inversion-Free Flow-based Image Editing
18. ICLR 2026Follow-Your-Shape: Shape-Aware Image Editing via Trajectory-Guided Region Control
19. ICLR 2026Free Lunch for Stabilizing Rectified Flow Inversion
20. ICLR 2026Geometric Image Editing via Effects-Sensitive In-Context Inpainting with Diffusion Transformers
21. ICLR 2026IC-Custom: Diverse Image Customization via In-Context Learning
22. ICLR 2026IVEBench: Modern Benchmark Suite for Instruction-Guided Video Editing Assessment
23. ICLR 2026LaTo: Landmark-tokenized Diffusion Transformer for Fine-grained Human Face Editing
24. ICLR 2026LazyDrag: Enabling Stable Drag-Based Editing on Multi-Modal Diffusion Transformers via Explicit Correspondence
25. ICLR 2026Learning an Image Editing Model without Image Editing Pairs
26. ICLR 2026Nano3D: A Training-Free Approach for Efficient 3D Editing Without Masks
27. ICLR 2026OmniText: A Training-Free Generalist for Controllable Text-Image Manipulation
28. ICLR 2026PICABench: How Far are We from Physical Realistic Image Editing?
29. ICLR 2026PICS: Pairwise Image Compositing with Spatial Interactions
30. ICLR 2026Pixel-Perfect Puppetry: Precision-Guided Enhancement for Face Image and Video Editing
31. ICLR 2026Positional Encoding Field
32. ICLR 2026Reconstruction Alignment Improves Unified Multimodal Models
33. ICLR 2026RegionE: Adaptive Region-Aware Generation for Efficient Image Editing
34. ICLR 2026TINKER: Diffusion's Gift to 3D--Multi-View Consistent Editing From Sparse Inputs without Per-Scene Optimization
35. ICLR 2026Training-Free Reward-Guided Image Editing via Trajectory Optimal Control
36. ICLR 2026Training-Free Text-Guided Color Editing with Multi-Modal Diffusion Transformer
37. ICLR 2026VINCIE: Unlocking In-context Image Editing from Video
38. ICLR 2026Visual Autoregressive Modeling for Instruction-Guided Image Editing
39. ICLR 2026W-EDIT: A Wavelet-Based Frequency-Aware Framework for Text-Driven Image Editing
40. ICLR 2026WorldEdit: Towards Open-World Image Editing with a Knowledge-Informed Benchmark
41. ICLR 20253DitScene: Editing Any Scene via Language-guided Disentangled Gaussian Splatting
42. ICLR 2025A3D: Does Diffusion Dream about 3D Alignment?
43. ICLR 2025ACE: All-round Creator and Editor Following Instructions via Diffusion Transformer
44. ICLR 2025Can Knowledge Editing Really Correct Hallucinations?
45. ICLR 2025CLIPDrag: Combining Text-based and Drag-based Instructions for Image Editing
46. ICLR 2025CollabEdit: Towards Non-destructive Collaborative Knowledge Editing
47. ICLR 2025Composable Interventions for Language Models
48. ICLR 2025ConceptPrune: Concept Editing in Diffusion Models via Skilled Neuron Pruning
49. ICLR 2025DiffusionGuard: A Robust Defense Against Malicious Diffusion-based Image Editing
50. ICLR 2025Discriminating image representations with principal distortions

常见问题

2023–2026 年四大会议一共有多少篇图像编辑论文？

共 692 篇，按年份分别是 2023 年 112 篇、2024 年 160 篇、2025 年 182 篇、2026 年 238 篇。

NeurIPS 2026 有多少篇图像编辑论文？

NeurIPS 2026 有 52 篇图像编辑论文，上一届是 45 篇。

ICLR 2026 有多少篇图像编辑论文？

ICLR 2026 有 40 篇图像编辑论文，上一届是 29 篇。

ICML 2026 有多少篇图像编辑论文？

ICML 2026 有 33 篇图像编辑论文，上一届是 11 篇。

CVPR 2026 有多少篇图像编辑论文？

CVPR 2026 有 113 篇图像编辑论文，上一届是 97 篇。

哪个会议的图像编辑论文最多？

CVPR，2023–2026 年共 364 篇。