Research map / 3D vision / 3D generation
DreamEdit3D: Personalization of Multi-View Diffusion Models for 3D Editing
Jinxin Ai, Matthias Niessner, Ziya Erkoç
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 |
|---|---|---|---|
| joint training on 4 orthogonal views instead of front view only (ablation) | DreamEdit3D · Robot Sitting case | GPT-4V overall: 5 -> 7 | Improved |
| joint training on 4 views instead of back view only (ablation) | DreamEdit3D · Robot Sitting case | GPT-4V overall: 3 -> 7 | Improved |
| joint training on 4 views instead of side view only (ablation) | DreamEdit3D · Robot Sitting case | CLIP dir-cos: -1.39 -> 3.15 | Improved |
| add cross attention loss (ablation) | DreamEdit3D · Robot Sitting case | CLIP dir: 0.11 -> 0.51 | Improved |
| add masked diffusion loss (ablation) | DreamEdit3D · Robot Sitting case | CLIP dir: -0.27 -> 0.51 | Improved |
| add both mask losses (ablation) | DreamEdit3D · Robot Sitting case | CLIP dir: -0.41 -> 0.51 | Improved |
| add DreamBooth fine-tuning to textual inversion (ablation) | DreamEdit3D · Robot Sitting case | GPT-4V identity: 1 -> 9 | Improved |
| add DreamBooth fine-tuning to textual inversion (ablation) | DreamEdit3D · Robot Sitting case | CLIP dir: 4.80 -> 0.51 | Hurt |
| add textual inversion before DreamBooth (ablation) | DreamEdit3D · Robot Sitting case | GPT-4V identity: 8 -> 9 | Improved |
| add textual inversion before DreamBooth (ablation) | DreamEdit3D · Robot Sitting case | CLIP dir: 0.40 -> 0.51 | 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.
- PrEditor3D — PrEditor3D: Fast and Precise 3D Shape Editing (CVPR 2025) · better on 6 of 6 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
- Omni-3DEdit: Generalized Versatile 3D Editing in One-Pass (CVPR 2026 · similarity 0.88)
- DreamCatalyst: Fast and High-Quality 3D Editing via Controlling Editability and Identity Preservation (ICLR 2025 · similarity 0.87)
- Progressive3D: Progressively Local Editing for Text-to-3D Content Creation with Complex Semantic Prompts (ICLR 2024 · similarity 0.87)
- InstructPix2NeRF: Instructed 3D Portrait Editing from a Single Image (ICLR 2024 · similarity 0.87)
- PrEditor3D: Fast and Precise 3D Shape Editing (CVPR 2025 · similarity 0.86)
- Energy-Guided Optimization for Personalized Image Editing with Pretrained Text-to-Image Diffusion Models (AAAI 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.
