Milestone for Social Learning Research Sample Clauses

Milestone for Social Learning Research. Social Learning Milestone I (~ next 2 years). The first target in social research involves studying and implementing non- verbal social cues for language and skill learning. The second target is modeling holophrase acquisition via intermodal learning; this entails sensitivity to aspects of acoustic packaging (cf. Sec. IV.C). The first target attempts to exploit biased learning via a form of rudimentary intentional reference. This can be achieved via joint attention between robot and human whereby the robot responds to gaze direction, mirroring and turn-taking in the interaction with the human interaction partner. The non-verbal clues direct robot attention to the actions or objects. Language acquisition proceeds by associating the robot's focus of attention (including its full sensorimotor feedback) with salient aspects of the human’s speech modality. The second challenge regards the modeling of holophrase acquisition via intermodal learning. This particularly refers to the implementation of the acoustic packaging that automatically permits the division of a sequence of events into units and thus there is synchrony between language and events. Social Learning Milestone II (~ next 4 years). The roadmap development in a 4-year perspective within the social learning scenarios expects that an ability to detect and exploit tutoring interactions will be developed in humanoid robots. This would be achieved by extending and enhancing the developments in previous milestones. Scaffolded learning of hierarchical behaviors in social interaction and the learning of grammar and vocabulary complement and enhance each other. Additionally further research on joint intentional framing and referential intent should be carried out together with the basic ideas for acquisition of negation usage of various types (e.g. refusal, absence, prohibition, propositional denial). Most of the latter require some modeling of motivation (volition and affect) on the part of the robot, as well as temporal scope encompassing memories and habits. Tutoring plays an important role in understanding actions. Research would consider how tutoring could be used for learning, how complex actions could be structured, which kind of units could be observed and how speech/sound signals (acoustic packaging) could be modeled. Studies would also be carried out to extend previous research in order to establish how to enhance rudimentary intentional reference to more sophisticated mechanisms for joint intentio...