Isn't this section incorrect?

charlesq34 / pointnet

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

Other

4.64k stars 1.44k forks source link

Isn't this section incorrect? #300

Open mathephysicist opened 1 year ago

mathephysicist commented 1 year ago

https://github.com/charlesq34/pointnet/blob/539db60eb63335ae00fe0da0c8e38c791c764d2b/models/transform_nets.py#L83

You set the weights and bias to zero. Then matmul.

Hence you will always have zero updates to the TNet that is prior.

jaidonlybbert commented 1 year ago

I believe you are correct. Also, the associated paper makes a leap here in the described output of the TNet. The (first) TNet should output a 3x3, however the final specified layer has an output dimension of 256. Unfortunate bit of ambiguity.