search

sjtuplayer / few-shot-diffusion

[ICCV 2023] Phasic Content Fusing Diffusion Model with Directional Distribution Consistency for Few-Shot Model Adaption
53 stars 2 forks source link

"generate source-domain images" during inference #5

Open fikry102 opened 10 months ago

fikry102 commented 10 months ago

If you hope to generate the target domain images, the simplest way is to generate source-domain images into $source_img_path, and then transfer these images to the target domain.

Can I copy some source-domain images from source dataset to $source_img_path (like "source_img_ffhq")? Or I need to use the pretrained diffusion model on the source domain to generate source-domain images first?

sjtuplayer commented 10 months ago

If you hope to generate the target domain images, the simplest way is to generate source-domain images into $source_img_path, and then transfer these images to the target domain.

Can I copy some source-domain images from source dataset to $source_img_path (like "source_img_ffhq")? Or I need to use the pretrained diffusion model on the source domain to generate source-domain images first?

You can just use the source-domain dataset as the inputs

15634960802 commented 5 months ago

我的结果只是源数据集的图像加上了噪声，生成不了目标域的图像啊？问题出在哪里啊