Abstract: We present Pow3R, a novel large 3D vision regression model that is highly versatile in the input modalities it accepts. Unlike previous feed-forward models that lack any mechanism to exploit ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results