search

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
5.08k stars 331 forks source link

About face datasets #217

Open y1xuan0726 opened 9 months ago

y1xuan0726 commented 9 months ago

About face datasets, I found that the Laion datasets is very large. I filtered some keywords based on captions (man/woman/person/child/...), but there are still most of the pictures that do not meet the expectations. Can you tell me how you selected the data? (Or just use a face recognition model to evaluate a very large dataset?)

xiaohu2015 commented 9 months ago

you can use laion-face dataset

in fact, I just used a face detection model to filter data...

y1xuan0726 commented 9 months ago

thank you so much !