PuLID: Pure and Lightning ID Customization via Contrastive Alignment
MP3•Episode hjem
Manage episode 440900318 series 2954468
Indhold leveret af Rob. Alt podcastindhold inklusive episoder, grafik og podcastbeskrivelser uploades og leveres direkte af Rob eller deres podcastplatformspartner. Hvis du mener, at nogen bruger dit ophavsretligt beskyttede værk uden din tilladelse, kan du følge processen beskrevet her https://da.player.fm/legal.
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 will be available at https://github.com/ToTheBeginning/PuLID
2024: Zinan Guo, Yanze Wu, Zhuowei Chen, Lang Chen, Qian He
https://arxiv.org/pdf/2404.16022v1
…
continue reading
2024: Zinan Guo, Yanze Wu, Zhuowei Chen, Lang Chen, Qian He
https://arxiv.org/pdf/2404.16022v1
299 episoder