Accessing T-Net in PointNet in Flux3D.jl

Specific Domains Machine Learning
1

I have a question about PointNet in Flux3D.jl.

In the original PointNet paper, the author trained the “T-net” to make PointNet invariant w.r.t. rigid transformations of input point clouds (red dashed box below):

Screen Shot 2022-09-13 at 16.44.33
Screen Shot 2022-09-13 at 16.44.331360×502 101 KB

Is there a way to access this trained T-net from PointNet of Flux3D.jl?