Abstract: Large Vision-Language Models (LVLMs) have shown impressive capabilities across various domains, but existing LVLMs have limited performance in dense perception and structured learning ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback