Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

PuLID: Pure and Lightning ID Customization via Contrastive Alignment

About

We propose Pure and Lightning ID customization (PuLID), a novel tuning-free ID customization method for text-to-image generation. By incorporating a Lightning T2I branch with a standard diffusion one, PuLID introduces both contrastive alignment loss and accurate ID loss, minimizing disruption to the original model and ensuring high ID fidelity. Experiments show that PuLID achieves superior performance in both ID fidelity and editability. Another attractive property of PuLID is that the image elements (e.g., background, lighting, composition, and style) before and after the ID insertion are kept as consistent as possible. Codes and models are available at https://github.com/ToTheBeginning/PuLID

Zinan Guo, Yanze Wu, Zhuowei Chen, Lang Chen, Peng Zhang, Qian He• 2024

Related benchmarks

TaskDatasetResultRank
Safe Generation RateI2P
GPT-4o Score82.47
9
Prompt-image AlignmentSneakyprompt
CLIPScore0.7211
8
Prompt-image AlignmentI2P
CLIPScore0.8195
8
Prompt-image AlignmentMMA-Diffusion
CLIPScore0.7228
8
Prompt-image AlignmentMisbinding
CLIPScore0.8722
8
Safe Generation RateSneakyprompt
GPT-4o0.7868
8
Safe Generation RateMMA-Diffusion
GPT-4o0.7146
8
Safe Generation RateMisbinding
GPT-4o Score0.6466
8
Identity-Preserving Text-to-Image GenerationIBench 41 prompts 100 IDs
Aesthetic Score68.3
7
Identity CustomizationIBench ChineseID editable long prompts
Aesthetic Score0.683
6
Showing 10 of 19 rows

Other info

Code

Follow for update