search

yzqin / s4g-release

S4G: Amodal Single-view Single-Shot SE(3) Grasp Detection in Cluttered Scene
35 stars 5 forks source link

PointNet2_tcl model pred #10

Closed xins981 closed 10 months ago

xins981 commented 10 months ago

@yzqin hi yuzhe. what means of the movable_logit?

yzqin commented 10 months ago

Hi @xins981 ,

The moveable_logit is actually something we tried but not used in the S4G paper and the experiments. It repressions the movement direction after grasping the object. However in the experiments we found that as long as our motion planner is strong enough, this information is not so useful. Thus, we do not use it in the S4G experiments.

You can just ignore this prediction.

xins981 commented 10 months ago

ok, thanks your explanation!