@jhawkins and @nleadholm summarize the brainstorming week in Milan. Mentioning details and issues related to behaviors: paired child behavior and morphology models, features in a behavior model are movements of the entire child, and removing state from morphology models. The team explored the issues and implications of removing the state and the mechanisms that could support this.
Summary Video
Main Video
0:00 Introduction
2:34 Details and Issues related to Behaviors
4:22 Behaviors Always Have a Child M & B Model
15:55 Features in a Behavior Model are Movements of the Entire Child
24:30 Paper: Gilbert 1977
32:48 How do we Represent Objects with Multiple Behaviors?
35:42 Exploring Morphology Models Without State
41:34 Locations in Morphology Models can be Associated with more than one Feature
52:23 The Changed Feature is Associatively Linked to a State of the Behavior Model
54:37 Open Questions and Issues
1:06:38 Another Related Proposal
1:23:21 Applying Motion to the Child Object
1:38:16 To Sum it Up