Research map / Image generation / Text-to-image diffusion
FastComposer: Tuning-Free Multi-subject Image Generation with Localized Attention
Guangxuan Xiao, Tianwei Yin, William T. Freeman, Frédo Durand, Song Han
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 cross-attention localization loss (ablation) | FastComposer · Multi-subject generation | identity preservation: 37.66% -> 43.11% | Improved |
| add cross-attention localization loss (ablation) | FastComposer · Multi-subject generation | prompt consistency: 25.03% -> 24.25% | Hurt |
| raise ratio of timesteps with subject conditioning (ablation) | FastComposer delayed conditioning · Single-subject | identity vs editability | Mixed |
| replace per-subject fine-tuning with tuning-free subject embedding | FastComposer vs Custom Diffusion · Single-subject | identity preservation: 43.37% -> 51.41% | Improved |
| replace fine-tuning with tuning-free subject embedding | FastComposer vs Custom Diffusion · Multi-subject | identity preservation: 5.37% -> 43.11% | Improved |
| replace fine-tuning with tuning-free subject embedding | FastComposer vs Textual-Inversion · Single-subject | total time: 2500 s -> 2 s | Improved |
Later papers that compared with this method
16 later accepted papers report FastComposer as a baseline in their results tables; 13 of them report a better number on most of the results they share with it.
- OmniPortrait: Fine-Grained Personalized Portrait Synthesis via Pivotal Optimization (ICLR 2026 · better on 4 of 4 shared results)
- ConsistentID: Portrait Generation With Multimodal Fine-Grained Identity Preserving (TPAMI 2026 · better on 4 of 4 shared results)
- FlowFace: Rectifying Identity Conditioning with Riemannian Geometry for Face Generation (ECCV 2026 · better on 3 of 3 shared results)
- PersonaHOI: Effortlessly Improving Face Personalization in Human-Object Interaction Generation (CVPR 2025 · better on 9 of 11 shared results)
- UniPortrait: A Unified Framework for Identity-Preserving Single- and Multi-Human Image Personalization (ICCV 2025 · better on 6 of 6 shared results)
- SerialGen: Personalized Image Generation by First Standardization Then Personalization (CVPR 2025 · better on 3 of 3 shared results)
- Foundation Cures Personalization: Improving Personalized Models’ Prompt Consistency via Hidden Foundation Knowledge (NeurIPS 2025 · better on 2 of 2 shared results)
- DynamicID: Zero-Shot Multi-ID Image Personalization with Flexible Facial Editability (ICCV 2025 · better on 2 of 2 shared results)
8 more not listed.
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
- MS-Diffusion: Multi-subject Zero-shot Image Personalization with Layout Guidance (ICLR 2025 · similarity 0.88)
- JeDi: Joint-Image Diffusion Models for Finetuning-Free Personalized Text-to-Image Generation (CVPR 2024 · similarity 0.87)
- DynASyn: Multi-Subject Personalization Enabling Dynamic Action Synthesis (AAAI 2025 · similarity 0.87)
- InstantBooth: Personalized Text-to-Image Generation without Test-Time Finetuning (CVPR 2024 · similarity 0.87)
- Large-Scale Text-to-Image Model with Inpainting is a Zero-Shot Subject-Driven Image Generator (CVPR 2025 · similarity 0.87)
- High-fidelity Person-centric Subject-to-Image Synthesis (CVPR 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.
