1 paper
Qingchuan Ma, Yuexiao Ma, Xiang Meng +7
Can vision-language models turn 2D observations into 3D understanding that supports precise action? This requires inferring object geometry, anticipating the effects of rotation, a…