Projects

Phase 2 Projects

Development as autonomous learning: Emergence of developmental stages that lead from sensori-motor behaviors to embodied cogniti


Project leaders: Gregor Schöner (Bochum)

Researchers: Yulia Sandamirskaya

Administration: Angelika Wille

Associates:

Summary:

Given the substantial individual differences between humans even in infancy, it is  amazing how infants and children go through stages of their early development in a reproducible order. Human development may be the most dramatic existence proof of autonomous learning: no designer sets the learning goals, most sensory information is self-generated, and much development is learning to learn. And yet, development proceeds in orderly fashion. The goal of this project is to provide specific theoretical proposals for how autonomous learning during development is self-organized. Within the neurally grounded framework of Dynamic Field Theory we will formulate models of autonomous learning and implement these in simulation and on autonomous robots. The behaviors of looking, perceiving, orienting in a scene, and reaching toward objects form the background on which Piaget’s conception was based that cognition emerges from the development of sensori-motor capacities. We will build our models in these specific behavioral contexts and assess them through qualitative comparison with experimental evidence and through experimental and theoretical analysis.

Phase 1 Projects

Development as autonomous learning: Emergence of developmental stages that lead from sensori-motor behaviors to embodied cogniti


Project leaders: Gregor Schöner (Bochum)

Researchers: Yulia Sandamirskaya

Administration: Angelika Wille

Associates:

Summary:

Given the substantial individual differences between humans even in infancy, it is  amazing how infants and children go through stages of their early development in a reproducible order. Human development may be the most dramatic existence proof of autonomous learning: no designer sets the learning goals, most sensory information is self-generated, and much development is learning to learn. And yet, development proceeds in orderly fashion. The goal of this project is to provide specific theoretical proposals for how autonomous learning during development is self-organized. Within the neurally grounded framework of Dynamic Field Theory we will formulate models of autonomous learning and implement these in simulation and on autonomous robots. The behaviors of looking, perceiving, orienting in a scene, and reaching toward objects form the background on which Piaget’s conception was based that cognition emerges from the development of sensori-motor capacities. We will build our models in these specific behavioral contexts and assess them through qualitative comparison with experimental evidence and through experimental and theoretical analysis.