Abstract: The development of large-scale image-text pair datasets has significantly advanced self-supervised learning in Vision-Language Processing (VLP). However, directly applying general-domain ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback