Tutorial on Embodiment

 

Welcome to the tutorial on embodiment


Editors:

Matej Hoffmann, Dorit Assaf, Rolf Pfeifer

Artificial Intelligence Laboratory, Department of Informatics
University of Zurich, Switzerland

Contact: Matej Hoffmann (matej [dot] hoffmann [at] iit.it)


Tutorial Content:

1. Introduction

1.1. Outline and Further Reading

1.2. Target Audience

2. Traditional Artificial Intelligence, its Successes and Problems

2.1. The Study of Intelligence

2.2. Successes of Traditional AI

2.3. Problems of Traditional AI

3. Embodiment

3.1. Theoretical Scheme

3.2. Locomotion Case Studies

3.2.1. Brain-less and power-less creatures

3.2.1.1. Euthanized trout

3.2.1.2. Passive dynamic walkers

3.2.1.3. Exercise

3.2.2. Brain-less and powered creatures

3.2.2.1. Behavioral diversity - crazy bird

3.2.2.2. Self-stabilization

3.2.2.3. Exercise

3.2.3. Simple control exploiting body dynamics and interaction with environment

3.2.3.1. Leg coordination in insect walking

3.2.3.2. Exploiting self-adaptation: cockroaches climbing over obstacles

3.2.3.3. Exercise

3.3. Grasping Case Studies

3.3.1. Cheap grasping with a robotic hand

3.3.2. Cheap grasping with an universal gripper

3.3.3. Exercise

3.4. Additional Online Lectures

4. Information Theoretic Implications of Embodiment

4.1. The Role of Sensory Morphology in Perception

4.1.1. Eye morphology

4.1.2. Slippage detection through skin morphology

4.2. Information Self-structuring through Sensory-motor Coordination

4.2.1. Visual perception case study

4.2.2. Quantifying information structure

4.3. Exercise

5. Embodied Cognition

5.1. Cognitive Science Paradigms

5.1.1. Cognitivism

5.1.2. Connectionism

5.1.3. Embodied dynamicism and enactivism

5.1.4. Summary

5.2. Learning Categories

5.2.1. Grounding toddler learning in sensory-motor dynamics

5.2.2. Categorization in artificial agents

5.2.3. Where mathematics comes from?

5.3. Cognition from Bottom-up

5.3.1. Body schema and forward models

5.3.2. Why is that cognition?

5.3.3. Cognitive development and the iCub humanoid robot

 

 

 

START TUTORIAL

Classical and modern view of intelligence.

Illustrations by Shun Iwasawa