Research map / Image generation / Text-to-image diffusion
WithAnyone: Toward Controllable and ID Consistent Image Generation
Hengyuan Xu, Wei Cheng, Peng Xing, Yixiao Fang, Shuhan Wu, Rui Wang, Xianfang Zeng, Daxin Jiang, Gang YU, Xingjun Ma, Yu-Gang Jiang
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 paired-data tuning phase (Phase 3) (ablation) | WithAnyone · MultiID-Bench 2-people subset | Copy-Paste (CP): 0.239 -> 0.161 | Improved |
| add paired-data tuning phase (Phase 3) (ablation) | WithAnyone · MultiID-Bench 2-people subset | Sim(GT): 0.406 -> 0.405 | No clear effect |
| add GT-aligned ID loss instead of self-aligned landmarks (ablation) | WithAnyone · MultiID-Bench 2-people subset | Sim(GT): 0.385 -> 0.405 | Improved |
| extend negative pool in InfoNCE from 63 in-batch to 4096 (ablation) | WithAnyone · MultiID-Bench 2-people subset | Sim(GT): 0.368 -> 0.405 | Improved |
| add extended negatives in InfoNCE (ablation) | WithAnyone · MultiID-Bench 2-people subset | Copy-Paste (CP): 0.074 -> 0.161 | Hurt |
| train on MultiID-2M instead of FFHQ only (ablation) | WithAnyone · MultiID-Bench 2-people subset | Sim(GT): 0.224 -> 0.405 | Improved |
| train with MultiID-2M and contrastive/paired training vs prior methods | WithAnyone · MultiID-Bench single-person subset | Copy-Paste (CP) at similar Sim(GT): InstantID 0.464 Sim(GT)/0.337 CP vs Ours 0.460/0.144 | Improved |
| use WithAnyone instead of best general models on OmniContext | WithAnyone · OmniContext single character subset | Overall: 6.52 vs OmniGen2 8.34 | Mixed |
| add paired tuning data (3-4 people) | WithAnyone · MultiID-Bench 3-and-4-people subset | Sim(GT): 0.379 -> 0.414 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- PuLID — PuLID: Pure and Lightning ID Customization via Contrastive Alignment (NeurIPS 2024) · better on 3 of 3 shared results
- ID-Patch — ID-Patch: Robust ID Association for Group Photo Personalization (CVPR 2025) · better on 2 of 2 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
- FaithfulFaces: Pose-Faithful Facial Identity Preservation for Text-to-Video Generation (NeurIPS 2026 · similarity 0.90)
- When StyleGAN Meets Stable Diffusion: a W+ Adapter for Personalized Image Generation (CVPR 2024 · similarity 0.89)
- Refine-by-Align: Reference-Guided Artifacts Refinement through Semantic Alignment (ICLR 2025 · similarity 0.88)
- DreamSalon: A Staged Diffusion Framework for Preserving Identity-Context in Editable Face Generation (CVPR 2024 · similarity 0.88)
- ARGUS: Stacked Multi-View Identity Mosaic Injection for Subject-Preserving Video Generation (NeurIPS 2026 · similarity 0.88)
- ID-Patch: Robust ID Association for Group Photo Personalization (CVPR 2025 · similarity 0.88)
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.
