Research map / Image generation / Editing and style transfer
Localized and attribute-specific editing
Edit specific objects and facial attributes while preserving background and conditions.
73 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2019–2026) · an approach in Editing and style transfer, keyword cluster “image editing · edits · inversion”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
18 papers in 2023–24 and 42 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 6.6% to 10.7%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2019 · CISI-net: Explicit Latent Content Inference and Imitated Style Rendering for Image Inpainting (AAAI 2019 · 20 citations)
- 2020 · Semantic Image Manipulation Using Scene Graphs (CVPR 2020 · 122 citations)
- 2021 · DeFLOCNet: Deep Image Editing via Flexible Low-Level Controls (CVPR 2021 · 33 citations)
- 2022 · HairCLIP: Design Your Hair by Text and Reference Image (CVPR 2022 · 105 citations)
- 2023 · MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image Editing (NeurIPS 2023 · 61 citations)
- 2024 · FreeDrag: Feature Dragging for Reliable Point-based Image Editing (CVPR 2024 · 20 citations)
- 2025 · Predicting the Original Appearance of Damaged Historical Documents (AAAI 2025 · 8 citations)
- 2026 · DragNeXt: Rethinking Drag-Based Image Editing (AAAI 2026 · 2 citations)
Earliest papers here
- CISI-net: Explicit Latent Content Inference and Imitated Style Rendering for Image Inpainting (AAAI 2019 · 20 citations)
- Semantic Image Manipulation Using Scene Graphs (CVPR 2020 · 122 citations)
- Fashion Editing With Adversarial Parsing Learning (CVPR 2020 · 87 citations)
What has been tried
47 papers here report 490 measured changes (a component added, removed or swapped, compared with a baseline): helped 77%, no clear effect 4%, hurt 17%, mixed 2%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Add adaptive template feature updating | Human face model · CCSD (×10) | 0.82 -> 0.35 | Helped | FreeDrag: Feature Dragging for Reliable Point-based Image Editing (CVPR 2024 · 20 citations) |
| Use ERLM-predicted region location instead of DiffEdit region location | DFMM-Spotlight expanded test set · FID | 12.92 -> 10.77 | Helped | TexFit: Text-Driven Fashion Image Editing with Diffusion Models (AAAI 2024 · 18 citations) |
| add the frequency stream to the RGB stream | CIMD-R · F1 | 0.330 -> 0.335 | Helped | A New Benchmark and Model for Challenging Image Manipulation Detection (AAAI 2024 · 17 citations) |
| replace reference U-Net with CLIP Encoder | Part composition · SSIM | 0.70 -> 0.66 | Hurt | Zero-shot Image Editing with Reference Imitation (NeurIPS 2024 · 17 citations) |
| Remove KV injection during texture sampling | Non-stationary textures · visual quality | Hurt | Generating Non-Stationary Textures using Self-Rectification (CVPR 2024 · 11 citations) | |
| add character perceptual loss (CPLoss) | HDR28K testing set · FID | 0.8416 -> 0.7499 | Helped | Predicting the Original Appearance of Damaged Historical Documents (AAAI 2025 · 8 citations) |
| add the multi-resolution field | InterHand2.6M, 7 unseen poses · PSNR | 27.11 -> 27.64 | Helped | OHTA: One-shot Hand Avatar via Data-driven Implicit Priors (CVPR 2024 · 8 citations) |
| Add diffusion-process optimization to facial-recognition perturbation (CVL -> CVL-D) | InstructPix2Pix edits · FR | 0.667 -> 0.380 | Helped | Edit Away and My Face Will not Stay: Personal Biometric Defense against Malicious Generative Editing (CVPR 2025 · 7 citations) |
| Replace counterfactual abduction with null-text inversion | Fashion image editing · preservation of source-image details | Hurt | Text-Driven Fashion Image Editing with Compositional Concept Learning and Counterfactual Abduction (CVPR 2025 · 7 citations) | |
| Add counterfactual fine-tuning to HVAE | ADNI · model performance | Hurt | Benchmarking Counterfactual Image Generation (NeurIPS 2024 · 5 citations) | |
| add superpixel segmentation | Noto Emoji · MSE | 0.00085 -> 0.00032 | Helped | Less is More: Efficient Image Vectorization with Adaptive Parameterization (CVPR 2025 · 4 citations) |
| Optimize the editing direction instead of using the naïve direction | Face pose rotation · yaw-direction correlation R² | 0.744 -> 0.846 | Helped | Edit One for All: Interactive Batch Image Editing (CVPR 2024 · 4 citations) |
What reviewers will expect
Datasets and baselines named in the 24 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Baselines
- InstructPix2Pix · 3 of 24 papers
- BrushNet · 2 of 24 papers
- MasaCtrl · 2 of 24 papers
- MagicBrush · 2 of 24 papers
- LIVE · 2 of 24 papers
- O&R · 2 of 24 papers
- PhotoGuard · 2 of 24 papers
- EditShield · 2 of 24 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 60 papers; from OpenAlex and the papers' affiliations).
- Adobe · 4 papers
- Shenzhen University · 3 papers
- Shanghai Jiao Tong University · 3 papers
- Microsoft · 3 papers
- McGill University · 2 papers
- Wuhan University · 2 papers
- Beihang University · 2 papers
- South China University of Technology · 2 papers
- Shanghai AI Laboratory · 2 papers
- Hong Kong University of Science and Technology · 2 papers
Use it
Download BibTeX of 33 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Instruction-based diffusion editing (151 papers)
- Inversion-based diffusion editing (113 papers)
- GAN inversion for editing (25 papers)
Elsewhere in the topic
- Texture and sketch synthesis (130 papers)
- Image-to-image translation (127 papers)
- Arbitrary neural style transfer (124 papers)
- Identity-preserving face manipulation (88 papers)
- Diffusion face generation and editing (61 papers)
- Virtual try-on (46 papers)
- Pose-guided person image synthesis (29 papers)
- Talking face generation (21 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).