Abstract: Document Understanding (DU) in long-contextual scenarios with complex layouts remains a significant challenge in vision-language research. Although Large Vision-Language Models (LVLMs) excel ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results