Abstract: Large Vision-Language Models (LVLMs) suffer from severe object hallucinations, leading them to frequently generate outputs that do not correspond to the image content, significantly reducing ...
Abstract: Understanding and interpreting a script is essential for effective acting. Existing visualization methods, however, primarily focus on general narrative comprehension and often neglect ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results