Extension of feature transformation for images as inputs

charlesq34 / pointnet

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation

Other

4.71k stars 1.44k forks source link

Extension of feature transformation for images as inputs #154

Open ameliajimenez opened 5 years ago

ameliajimenez commented 5 years ago

Hello,

Great work! I came across it and I've seen that the function "feature_transform_net" is defined for inputs of size BxNx1xK.

Would it be possible to change this function for an input being: BxNxNxK? This type of inputs are common when working with images, instead of 3D points.

Thanks in advance!