Research map / Image generation / Text-to-image diffusion
Foundation Encoders Are All You Need for Preference-Aware Personalization
Hyungjin Kim, Seokho Ahn, Young-Duk Seo
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 |
|---|---|---|---|
| use target query as personalized query instead of unconditional query (ablation) | FAN · PIP | target CLIP score: 22.23 -> 25.47 | Improved |
| use target query as personalized query instead of unconditional query (ablation) | FAN · ML | target CLIP score: 24.63 -> 30.69 | Improved |
| add tailored profiling (ablation) | FAN · PIP | target CLIP score: 25.58 -> 25.47 | No clear effect |
| add PA skip (skip personalized attention in selected layers) (ablation) | FAN · PIP | target CLIP score: 24.98 -> 25.47 | Improved |
| add PA skip (skip personalized attention in selected layers) (ablation) | FAN · ML | history CLIP score: 12.94 -> 12.83 | Hurt |
| use tailored profiling instead of BM25 profiling (ablation) | FAN · PIP average CLIP score, 10% sampling | CLIP score: 17.71 -> 20.36 | Improved |
| use tailored profiling instead of random sampling (ablation) | FAN · PIP average CLIP score, 10% sampling | CLIP score: 20.30 -> 20.36 | No clear effect |
| use tailored profiling instead of coreset sampling (ablation) | FAN · ML average text align, 10% sampling | Text align: 67.88 -> 67.79 | No clear effect |
| apply FAN to Stable Diffusion V3 on ML | Stable Diffusion V3 · ML | CLIP score improvement rate: +2.22% | Improved |
| apply FAN on SD V1 ML versus unpersonalized model (target CLIP) | Stable Diffusion V1 · ML | target CLIP score: 31.27 -> 30.69 | Hurt |
| use FAN versus DrUM for text alignment | OpenCLIP ViT-L · PIP | target text align: 96.42 -> 95.94 | Hurt |
Closest papers by meaning
- Personalized Preference Fine-tuning of Diffusion Models (CVPR 2025 · similarity 0.85)
- Personalized Image Editing in Text-to-Image Diffusion Models via Collaborative Direct Preference Optimization (NeurIPS 2025 · similarity 0.85)
- Foundation Cures Personalization: Improving Personalized Models’ Prompt Consistency via Hidden Foundation Knowledge (NeurIPS 2025 · similarity 0.85)
- JeDi: Joint-Image Diffusion Models for Finetuning-Free Personalized Text-to-Image Generation (CVPR 2024 · similarity 0.84)
- Towards Unified Multi-Modal Personalization: Large Vision-Language Models for Generative Recommendation and Beyond (ICLR 2024 · similarity 0.84)
- InstantBooth: Personalized Text-to-Image Generation without Test-Time Finetuning (CVPR 2024 · similarity 0.84)
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.
