Research map / Image generation / Editing and style transfer
EraseDraw : Learning to Insert Objects by Erasing Them from Images
Alper Canberk, Maksym Bondarenko, Ege Ozguroglu, Ruoshi Liu, Carl Vondrick
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| learn insertion from erased-object data instead of InstructPix2Pix training data | EraseDraw · EmuEdit add-instruction validation | CLIPdir: 0.0976 -> 0.1383 | Improved |
| use EraseDraw instead of MagicBrush++ for insertion | EmuEdit add-instruction validation · EmuEdit add-instruction validation | human win rate (%): 55.06 -> 82.22 | Improved |
| mix autonomously generated EraseDraw-15k with GQA-Insert for training (ablation) | EraseDraw · EmuEdit add-instruction validation | CLIPdir: 0.12788 -> 0.1383 | Improved |
| train on 15,000 autonomously generated images instead of 50,000 GQA-derived images (ablation) | EraseDraw · EmuEdit add-instruction validation | CLIPout: 0.0907 -> 0.0905 | No clear effect |
| insert objects iteratively step by step instead of one-shot prompting with Stable Diffusion 1.5 | EraseDraw · iterative scene composition (qualitative) | prompt adherence | Improved |
| add CLIP-score beam search to select candidates across insertion steps | EraseDraw iterative generation · iterative scene composition (qualitative) | prompt adherence | Improved |
Later papers that compared with this method
1 later accepted paper reports EraseDraw as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Add-it: Training-Free Object Insertion in Images With Pretrained Diffusion Models (ICLR 2025 · better on 3 of 4 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- InstructPix2Pix — InstructPix2Pix: Learning To Follow Image Editing Instructions (CVPR 2023) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Paint by Inpaint: Learning to Add Image Objects by Removing Them First (CVPR 2025 · similarity 0.91)
- Add-it: Training-Free Object Insertion in Images With Pretrained Diffusion Models (ICLR 2025 · similarity 0.88)
- CLIPAway: Harmonizing focused embeddings for removing objects via diffusion models (NeurIPS 2024 · similarity 0.87)
- Attentive Eraser: Unleashing Diffusion Model’s Object Removal Potential via Self-Attention Redirection Guidance (AAAI 2025 · similarity 0.87)
- Towards Enhanced Image Inpainting: Mitigating Unwanted Object Insertion and Preserving Color Consistency (CVPR 2025 · similarity 0.87)
- Object-Aware Inversion and Reassembly for Image Editing (ICLR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
