Generalizable Pose Estimation Using Implicit Scene Representations

Understand

6-DoF pose estimation is an essential component of robotic manipulation pipelines.

• However, it usually suffers from a lack of generalization to new instances and object types.
• Most widely used methods learn to infer the object pose in a discriminative setup where the model filters useful information to infer the exact pose of the object.
• While such methods offer accurate poses, the model does not store enough information to generalize to new objects.