A complete end-to-end demonstration in which we collect training data in Unity and use that data to train a deep neural network to predict the pose of a cube. This model is then deployed in a simulated robotic pick-and-place task.
As for anyone trying to run the generated Message classes, this will not work with the current TCP-endpoint package (Version 0.7 and above )
here's a quick fix for the
The type or namespace name 'Pose' does not exist in the namespace 'RosMessageTypes.Geometry' (are you missing an assembly reference?)
or any namespace name errors
just Add Msg at the end of the message's class name
ex: public Geometry.Pose pick_pose;
to
public Geometry.PoseMsg pick_pose;
As for anyone trying to run the generated Message classes, this will not work with the current TCP-endpoint package (Version 0.7 and above ) here's a quick fix for the
or any namespace name errors
Msg
at the end of the message's class name ex:public Geometry.Pose pick_pose;
topublic Geometry.PoseMsg pick_pose;