Research map / Image generation / Text-to-image diffusion
FreeCus: Free Lunch Subject-driven Customization in Diffusion Transformers
Yanbing Zhang, Zhe Wang, Qin Zhou, Mengping Yang
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 pivotal attention sharing (PAS) (ablation) | FreeCus · DreamBench++ | DINO: 0.590 -> 0.696 | Improved |
| add PAS (ablation) | FreeCus · DreamBench++ | CLIP-I: 0.810 -> 0.853 | Improved |
| add PAS (ablation) | FreeCus · DreamBench++ | CLIP-T: 0.327 -> 0.308 | Hurt |
| add adjustment of noise shifting (ANS) (ablation) | FreeCus · DreamBench++ | DINO: 0.624 -> 0.696 | Improved |
| add ANS (ablation) | FreeCus · DreamBench++ | CLIP-I: 0.829 -> 0.853 | Improved |
| add semantic features compensation (SFC) (ablation) | FreeCus · DreamBench++ | DINO: 0.633 -> 0.696 | Improved |
| add SFC (ablation) | FreeCus · DreamBench++ | CLIP-I: 0.822 -> 0.853 | Improved |
| raise lambda_p and lambda_r from 1.00 to 1.10 (ablation) | FreeCus · DreamBench++ | DINO: 0.626 -> 0.696 | Mixed |
| increase negative shift magnitude from mu*0 to mu*-1.0 (ablation) | FreeCus · DreamBench++ | DINO: 0.648 -> 0.696 | Mixed |
| filter LVLM captions with LLM instead of unfiltered LVLM caption (ablation) | FreeCus · DreamBench++ | CLIP-T: 0.303 -> 0.308 | Improved |
| use detailed filtered captions instead of concise filtered captions (ablation) | FreeCus · DreamBench++ | DINO: 0.696 -> 0.682 | Hurt |
| combine FreeCus with Qwen2VL-Flux | Qwen2VL-Flux · DreamBench++ | CLIP-I: 0.841 -> 0.853 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- MS-Diffusion — MS-Diffusion: Multi-subject Zero-shot Image Personalization with Layout Guidance (ICLR 2025) · better on 0 of 3 shared results
- OminiControl — OminiControl: Minimal and Universal Control for Diffusion Transformer (ICCV 2025) · better on 2 of 3 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
- Rare Text Semantics Were Always There in Your Diffusion Transformer (NeurIPS 2025 · similarity 0.89)
- Prompt-Free Diffusion: Taking "Text" out of Text-to-Image Diffusion Models (CVPR 2024 · similarity 0.89)
- Exploring the Deep Fusion of Large Language Models and Diffusion Transformers for Text-to-Image Synthesis (CVPR 2025 · similarity 0.89)
- R&B: Region and Boundary Aware Zero-shot Grounded Text-to-image Generation (ICLR 2024 · similarity 0.88)
- MS-Diffusion: Multi-subject Zero-shot Image Personalization with Layout Guidance (ICLR 2025 · similarity 0.87)
- DreamDistribution: Learning Prompt Distribution for Diverse In-distribution Generation (ICLR 2025 · 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.
