Research map / Image generation / Editing and style transfer
VMDiff: Visual Mixing Diffusion for Limitless Cross-Object Synthesis
Zeren Xiong, Yue Yu, Ze-dong Zhang, Shuo Chen, Jian Yang, Jun Li
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 |
|---|---|---|---|
| add blending-noise refinement (BNoise) before mixing denoise (ablation) | VMDiff · IIOF | similarity score SS: 1.570 -> 1.586 | Improved |
| add alpha search (golden section) over mixing ratio (ablation) | VMDiff · IIOF | similarity score SS: 1.586 -> 2.025 | Improved |
| add beta1/beta2 search (ablation) | VMDiff · IIOF | similarity score SS: 2.025 -> 2.068 | Improved |
| interpolate embeddings instead of concatenating them in BNoise (ablation) | VMDiff · IIOF | similarity score SS: 1.586 -> 1.551 (before), 1.532 (after) | Hurt |
| use VMDiff vs best multi-concept baseline MIP-Adapter | MIP-Adapter · IIOF | similarity score SS: 1.866 -> 2.068 | Improved |
| VMDiff versus MIP-Adapter on LLaVA coherence | MIP-Adapter · IIOF | VQAScore-LLaVA single-entity coherence: 0.417 vs 0.413 | Hurt |
| human preference for VMDiff over baselines | GPT-4o / DreamO etc. · user study | vote share (%): 12.9 -> 67.3 | Improved |
| VMDiff iterative optimization cost | VMDiff · runtime per pair | time: 2 min 46 sec vs 8 sec (DreamO) | Hurt |
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.
- OmniGen — OmniGen: Unified Image Generation (CVPR 2025) · better on 4 of 4 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
- IV-mixed Sampler: Leveraging Image Diffusion Models for Enhanced Video Synthesis (ICLR 2025 · similarity 0.89)
- ViVid-1-to-3: Novel View Synthesis with Video Diffusion Models (CVPR 2024 · similarity 0.88)
- Novel Object Synthesis via Adaptive Text-Image Harmony (NeurIPS 2024 · similarity 0.87)
- NVS-Solver: Video Diffusion Model as Zero-Shot Novel View Synthesizer (ICLR 2025 · similarity 0.87)
- ScaleCrafter: Tuning-free Higher-Resolution Visual Generation with Diffusion Models (ICLR 2024 · similarity 0.87)
- Efficient-3Dim: Learning a Generalizable Single-image Novel-view Synthesizer in One Day (ICLR 2024 · similarity 0.87)
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.
