Abstract: Vision-language modeling (VLM) aims to bridge the information gap between images and natural language. Under the new paradigm of first pretraining on massive image-text pairs and then ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback