Abstract: Multimodal language models (MLMs) still face challenges in fundamental visual perception tasks where specialized models excel. Tasks requiring reasoning about 3D structures benefit from ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback