how to combine image prompts and conditions such as structural control?

tencent-ailab / IP-Adapter

The image prompt adapter is designed to enable a pretrained text-to-image diffusion model to generate images with image prompt.

Apache License 2.0

5.01k stars 326 forks source link

Closed ruaruaruabick closed 1 year ago

ruaruaruabick commented 1 year ago

take T2I-adapter as example, the inputs include a image prompt "p" and a condition "c", what is the input of unet?

xiaohu2015 commented 1 year ago

@ruaruaruabick hi, the structural control models are used same as their original implementation