SHI-Labs / Prompt-Free-Diffusion

Prompt-Free Diffusion: Taking "Text" out of Text-to-Image Diffusion Models, arxiv 2023 / CVPR 2024

Home Page:https://arxiv.org/abs/2305.16223

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Inpainting results

jgedev opened this issue · comments

commented

Curious if the authors or anyone else has attempted experiments substituting SeeCoder into an inpainting pipeline with the runwayml checkpoint or the ControlNet 1.1 inpaint preprocessor to achieve an effect similar to Paint by Example: https://arxiv.org/abs/2211.13227

WechatIMG54
WechatIMG55
这是我自己模型训练的结果,类似于Adobe的生成填充,微信最新的<换影CP>小程序,根据一张照片,一个模版,进行一键换头,一键换脸型。
不过需要不小的算力进行训练,我可能没有大量资源进行训练优化了,希望可以找到一些朋友进行合作研究和开发