FaRL offers powerful pre-training transformer backbones for face analysis tasks. Its pre-training combines both the image-text contrastive learning and the ...
確定! 回上一頁