MIRROR Project Review IST-2000-28159 Brussels – December 1 st, 2003.

Slides:



Advertisements
Similar presentations
Giacomo Rizzolatti and Corrado Sinigaglia. Basic knowledge Mirror mechanism Unifies perception and action Its functional role depends on its anatomical.
Advertisements

Imagery and observational learning Some (more) observations.
NATURE REVIEWS | NEUROSCIENCE SEP 01
Perception and Perspective in Robotics Paul Fitzpatrick MIT Computer Science and Artificial Intelligence Laboratory Humanoid Robotics Group Goal To build.
C534 Lecture 8: Action observation and the human ‘mirror system’
Disadvantages of Spherical/Polar Revolute Coordinate System.
Discoveries of Infancy
Yiannis Demiris and Anthony Dearden By James Gilbert.
SA-1 Body Scheme Learning Through Self-Perception Jürgen Sturm, Christian Plagemann, Wolfram Burgard.
Laboratory for Perceptual Robotics – Department of Computer Science Hierarchical Mechanisms for Robot Programming Shiraj Sen Stephen Hart Rod Grupen Laboratory.
L ABORATORY FOR P ERCEPTUAL R OBOTICS U NIVERSITY OF M ASSACHUSETTS A MHERST D EPARTMENT OF C OMPUTER S CIENCE A Relational Representation for Procedural.
Introduction to Robotics
Matching brain and body dynamics Daniel Wolpert: – "Why don't plants have brains?" – "Plants don't have to move!" Early phases of embodied artificial intelligence:
Laboratory for Perceptual Robotics Department of Computer Science University of Massachusetts Amherst Natural Task Decomposition with Intrinsic Potential.
Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme Neural Robot Control Cornelius Weber Hybrid Intelligent.
L ABORATORY FOR P ERCEPTUAL R OBOTICS U NIVERSITY OF M ASSACHUSETTS A MHERST D EPARTMENT OF C OMPUTER S CIENCE Intent Recognition as a Basis for Imitation.
Sensory-Motor Primitives as a Basis for Imitation: Linking Perception to Action and Biology to Robotics Presentation by Dan Hartmann 21 Feb 2006.
ADAPT IST Artificial Development Approach to Presence Technologies 2 nd Review Meeting Munich, June 7-9 th, 2004.
ICDL 2004 IST RobotCub RobotCub IST
 For many years human being has been trying to recreate the complex mechanisms that human body forms & to copy or imitate human systems  As a result.
Metta, Sandini, Natale & Panerai.  Developmental principles based on biological systems.  Time-variant machine learning.  Focus on humanoid robots.
IST – OMNIVIEWS Omnidirectional Visual System Final Review September 27-28, 2001 Lisbon Final Review September 27-28, 2001 Lisbon.
intelligence study and design of intelligent agentsis the intelligence of machines and the branch of computer science that aims to create it. AI textbooks.
Mind RACES from Reactive to Anticipatory Cognitive Embodied Systems Our Objectives The general goal of Mind RACES is to investigate different anticipatory.
INTEGRATED SYSTEMS 1205 Technology Education A Curriculum Review Sabine Schnepf-Comeau July 19, 2011 ED 4752.
Chapter 2 Robot Kinematics: Position Analysis
IST – OMNIVIEWS Omnidirectional Visual System Final Review September 27-28, 2001 Lisbon Final Review September 27-28, 2001 Lisbon.
Towards Cognitive Robotics Biointelligence Laboratory School of Computer Science and Engineering Seoul National University Christian.
The iCub humanoid robot: An open-systems platform for research in cognitive development Giorgio Metta et al., Neural networks 2010 윤성재.
Vision-Based Reach-To-Grasp Movements From the Human Example to an Autonomous Robotic System Alexa Hauck.
Beyond Gazing, Pointing, and Reaching A Survey of Developmental Robotics Authors: Max Lungarella, Giorgio Metta.
SEMINAR REPORT ON K.SWATHI. INTRODUCTION Any automatically operated machine that functions in human like manner Any automatically operated machine that.
Robotics Sharif In the name of Allah. Robotics Sharif Introduction to Robotics o Leila Sharif o o Lecture #2: The.
Vision-based human motion analysis: An overview Computer Vision and Image Understanding(2007)
Project ArteSImit Artefact Structural Learning through Imitation (TU München, U Parma, U Tübingen, U Minho, KU Nijmegen) Giorgio Panin - TUM.
What is it?  Development can be summed up as the changes we go through in life, not just physical changes.  Growth refers to the increase in cell number,
Visuo-Motor Relationships: Plasticity and Development Read: Rosenbaum Chapters 2, 6 + may.doc.
Natural Tasking of Robots Based on Human Interaction Cues Brian Scassellati, Bryan Adams, Aaron Edsinger, Matthew Marjanovic MIT Artificial Intelligence.
Cognitive Science Overview Cognitive Science Defined The Brain Assumptions of Cognitive Science Cognitive Information Processing Cognitive Science and.
Object Lesson: Discovering and Learning to Recognize Objects Object Lesson: Discovering and Learning to Recognize Objects – Paul Fitzpatrick – MIT CSAIL.
Week 2-1: Human Information Processing
Development 1: Cognitive Development Josée L. Jarry, Ph.D., C.Psych. Introduction to Psychology Department of Psychology University of Toronto July 7,
Chapter 7. Learning through Imitation and Exploration: Towards Humanoid Robots that Learn from Humans in Creating Brain-like Intelligence. Course: Robots.
ROBOT VISION LABORATORY 김 형 석 Robot Applications-B
Chapter 10. The Explorer System in Cognitive Systems, Christensen et al. Course: Robots Learning from Humans On, Kyoung-Woon Biointelligence Laboratory.
Chapter 8. Learning of Gestures by Imitation in a Humanoid Robot in Imitation and Social Learning in Robots, Calinon and Billard. Course: Robots Learning.
INSTITUTO DE SISTEMAS E ROBÓTICA Computer and Robot Vision Laboratory MIRROR DESIGN: Principles and Tools Cláudia.
Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme IST ~ INPRO (Information Processing by Natural.
Giorgio Metta · Paul Fitzpatrick Humanoid Robotics Group MIT AI Lab Better Vision through Manipulation Manipulation Manipulation.
Investigating the basis for conversation between human and robot Experiments using natural, spontaneous speech, speaking to the robot as if it were a small.
Cognitive Science Overview Introduction, Syllabus
Visuo-Motor Relationships: Plasticity and Development.
Intelligent Control Methods Lecture 2: Artificial Intelligence Slovak University of Technology Faculty of Material Science and Technology in Trnava.
COMPUTER INTEGRATED MANUFACTURING Evgeny Kagan CIM Laboratory, room 454.
1 Azhari, Dr Computer Science UGM. Human brain is a densely interconnected network of approximately neurons, each connected to, on average, 10 4.
Ali Ghadirzadeh, Atsuto Maki, Mårten Björkman Sept 28- Oct Hamburg Germany Presented by Jen-Fang Chang 1.
Functionality of objects through observation and Interaction Ruzena Bajcsy based on Luca Bogoni’s Ph.D thesis April 2016.
San Diego May 22, 2013 Giovanni Saponaro Giampiero Salvi
Modeling the development of mirror neurons Problem Solution
Computer Output Device: Arm Robot
A. Melendez-Calderon, L. Masia, R. Gassert, G. Sandini, E. Burdet
Domo: Manipulation for Partner Robots Aaron Edsinger MIT Computer Science and Artificial Intelligence Laboratory Humanoid Robotics Group
Manipulation in Human Environments
CHAPTER 2 FORWARD KINEMATIC 1.
Opportunistic Perception
Disadvantages of Spherical/Polar
Social neuroscience Domina Petric, MD.
Better Vision through Experimental Manipulation
The Organization and Planning of Movement Ch
Domo: Manipulation for Partner Robots Aaron Edsinger MIT Computer Science and Artificial Intelligence Laboratory Humanoid Robotics Group
Presentation transcript:

MIRROR Project Review IST Brussels – December 1 st, 2003

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR AgendaAgenda 1. Overview of Mirror Projects – Giorgio Metta First Hour Introduction 2. Model of the learning artifact – Jose’ Santos-Victor 3. Biological experiments – Luciano Fadiga 4. Development of action – Claes von Hofsten 2. Model of the learning artifact – Jose’ Santos-Victor 3. Biological experiments – Luciano Fadiga 4. Development of action – Claes von Hofsten Highlights of the second year results

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR ConsortiumConsortium LIRA-Lab, DIST, University of Genova – Italy: G. Sandini, G. Metta Dept. of Biomedical Sciences University of Ferrara – Italy: Luciano Fadiga, Laila Craighero Instituto Superior Técnico – Lisbon, Portugal: Josè Santos-Victor, Alex Bernardino, Manuel Cabido-Lopes Dept. of Psychology – University of Uppsala – Sweden: Claes Von Hofsten, Kerstin Rosander

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Scientific framework The project investigates the association between visual information and motor commands in the learning, representation and understanding of complex manipulative gestures (grasping). taking inspiration from…

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Mirror neurons (and ‘canonical’ neurons) The neuron is activated by “seeing” someone else’s hand performing a manipulative action and while the monkey is performing the same action A mirror neuron is a motor neuron that “recognizes” gestures

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Projects objectives n Realize an artificial system (artifact) able to interpret human gestures (hand grasping) by means of a mirror system n Study the ontogenesis of motor (mirror) representation in humans, monkeys, and robots n Realize an artificial system (artifact) able to interpret human gestures (hand grasping) by means of a mirror system n Study the ontogenesis of motor (mirror) representation in humans, monkeys, and robots MIRROR combines physiology, psychology, and computer science in building a model of the mirror system. In doing so, we aim at enlarging our understanding of action recognition, interpretation, and communication in biological systems. MIRROR combines physiology, psychology, and computer science in building a model of the mirror system. In doing so, we aim at enlarging our understanding of action recognition, interpretation, and communication in biological systems.

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Developmental approach …understanding someone else’s actions is much simpler by being able of performing that action In other words the person’s ability in interpreting and/or imitating motor acts is facilitated by having learned how to perform that action while observing, visually and “motorically”, his/her own body

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Elements of the model n What do two individuals have in common? –The effects of the action are shared (visually for example) –The target object of the action maybe shared –A similar (if not the same) motor repertoire n What do two individuals have in common? –The effects of the action are shared (visually for example) –The target object of the action maybe shared –A similar (if not the same) motor repertoire

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Conceptual schema mov Action B mov Action C Perceivable consequences Action A Sensory feedback (GOAL) C

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Exchange of information Individual AIndividual B A A B B

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Object affordances n An individual can apply his/her own experience in determining which actions are more likely to be successful in manipulating an object n Knowledge about objects can be acquired by actively manipulating them n An individual can apply his/her own experience in determining which actions are more likely to be successful in manipulating an object n Knowledge about objects can be acquired by actively manipulating them

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Experimenting w/ affordances Bottle, “pointiness”=0.13Car, “pointiness”=0.07 Ball, “pointiness”=0.02Cube, “pointiness”=0.03 estimated probability of occurrence difference between angle of motion and principal axis of object [degrees] Rolls at right angles to principal axis Rolls along principal axis

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Development of mirror system Interpretation of other individual’s actions can be acquired gradually… n Learn how to control the hand (acquire motor actions and their visual consequences) n Learn to associate feasible grasping to object characteristics (associate between objects with their affordances) n Learn to associate actions performed by others to our own internal representation (recognize)  mirror system Interpretation of other individual’s actions can be acquired gradually… n Learn how to control the hand (acquire motor actions and their visual consequences) n Learn to associate feasible grasping to object characteristics (associate between objects with their affordances) n Learn to associate actions performed by others to our own internal representation (recognize)  mirror system

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR What we are going to show n A formal probabilistic model of affordances and recognition based on motor information n Investigation on the role of visual information for the response of mirror neurons n Investigation on the contribution of visual information in interpreting someone else’s action n Investigation on the development of action in human infants …and of course results! n A formal probabilistic model of affordances and recognition based on motor information n Investigation on the role of visual information for the response of mirror neurons n Investigation on the contribution of visual information in interpreting someone else’s action n Investigation on the development of action in human infants …and of course results!

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR …what’s next? n Full fledged implementation on a robotic system n Better understanding of the model ergo… n Better understanding of biological systems n Full fledged implementation on a robotic system n Better understanding of the model ergo… n Better understanding of biological systems

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Intentionally blank

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Development…Development… n Reaching and grasping (1 or 2 mechanisms?) UU – Rotating rod (Babies) DIST – Posting Task (Robot) n Understand effects of manipulation actions n Reaching and grasping (1 or 2 mechanisms?) UU – Rotating rod (Babies) DIST – Posting Task (Robot) n Understand effects of manipulation actions

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Two sensorimotor components n Visual: how a grasping action looks visually (object + hand) n Motor/proprioceptive (the motor commands to generate a grasping and the resulting proprioceptive signals) n Visual: how a grasping action looks visually (object + hand) n Motor/proprioceptive (the motor commands to generate a grasping and the resulting proprioceptive signals) Performed grasps provide all components while observed grasp are only visual Performed grasps provide all components while observed grasp are only visual A grasping action has two components

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR The role of vision n How can the association between motor, proprioceptive and visual components be exploited during learning self-generated actions? n What is the role of vision in learning how to grasps? n Which visual information is required to “recognize”? n How can the association between motor, proprioceptive and visual components be exploited during learning self-generated actions? n What is the role of vision in learning how to grasps? n Which visual information is required to “recognize”?

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Visuo-motor association n Setup for the acquisition of visual and proprioceptive information to test different learning strategies and visual and motor parameters DIST and IST

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Grasp Data Acquisition Stereo vision Kinematic data

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Grasp Data Acquisition good matching!! cylindrical Power grasp derivatives: more noisy, but similar (bell shaped!)

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Grasp-related Visual Data n Early processing of grasp-related visual data (Stereo – viewpoint transformation) Viewpoint Transformation Stereo Vision IST

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Role of vision in Mirror System n Test the influence of “seeing the hand” and the object in the mirror system (Setup) UNIFE and IST

Project funded by the Future and Emerging Technologies arm of the IST Programme FET-Open scheme “Artefefacts that live and grow” IST MIRROR Robot Hand n 16 degrees joints n 6 controlled d.o.f. n Elastic components on all joints n Hall sensors on all joints