Abstract: Vision-Language Pre-training (VLP) that utilizes the multi-modal information to promote the training efficiency and effectiveness, has achieved great success in vision recognition of natural ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results