Research map / Image generation / Editing and style transfer
ObjectDrop: Bootstrapping Counterfactuals for Photorealistic Object Removal and Insertion
Daniel Winter, Matan Cohen, Shlomi Fruchter, Yael Pritch, Alex Rav-Acha, Yedid Hoshen
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 |
|---|---|---|---|
| Fine-tune on counterfactual before/after pairs instead of using a pre-trained inpainting model | Inpainting diffusion model · Object removal, held-out test set | PSNR: 21.192 -> 23.153 | Improved |
| Add bootstrap supervision (pre-train on 350K synthetic counterfactuals) before fine-tuning on real pairs (ablation) | ObjectDrop insertion model · Object insertion, held-out test set | PSNR: 20.178 -> 21.625 | Improved |
| Add bootstrap supervision (ablation) | ObjectDrop insertion model · Object insertion, held-out test set | LPIPS: 0.066 -> 0.057 | Improved |
| Increase counterfactual dataset size for removal (ablation) | ObjectDrop removal model · Object removal | visual quality | Improved |
| Pre-train from text-to-image model instead of inpainting model (ablation) | ObjectDrop removal · Object removal, shadows and reflections | visual quality | No clear effect |
| Apply counterfactual fine-tuning on public Stable-Diffusion-Inpainting v1 | SD Inpainting v1 · Object removal, held-out test set | PSNR: 19.198 -> 21.363 | Improved |
| Train insertion on the 2,500-pair real dataset only (ablation) | ObjectDrop insertion · Object insertion | visual quality | Hurt |
| Use ObjectDrop instead of Emu Edit and MGIE for removal | ObjectDrop · Emu Edit benchmark user study | user preference: 35.9% -> 64.1% (Emu Edit); 13.5% -> 86.5% (MGIE) | Improved |
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.
- Emu Edit — Emu Edit: Precise Image Editing via Recognition and Generation Tasks (CVPR 2024) · better on 1 of 1 shared result
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
- Diffusion Handles Enabling 3D Edits for Diffusion Models by Lifting Activations to 3D (CVPR 2024 · similarity 0.87)
- PAIR Diffusion: A Comprehensive Multimodal Object-Level Image Editor (CVPR 2024 · similarity 0.87)
- Paint by Inpaint: Learning to Add Image Objects by Removing Them First (CVPR 2025 · similarity 0.87)
- PS-Diffusion: Photorealistic Subject-Driven Image Editing with Disentangled Control and Attention (CVPR 2025 · similarity 0.87)
- PixelMan: Consistent Object Editing with Diffusion Models via Pixel Manipulation and Generation (AAAI 2025 · similarity 0.86)
- ROSE: Remove Objects with Side Effects in Videos (NeurIPS 2025 · 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.
