Robot Programming by Demonstration

1. Auflage 2009
ISBN: 978-1-4398-0867-2
Verlag: CRC Press

Buch, Englisch, 320 Seiten, Gewicht: 612 g

Also referred to as learning by imitation, tutelage, or apprenticeship learning, Programming by Demonstration (PbD) develops methods by which new skills can be transmitted to a robot. This book examines methods by which robots learn new skills through human guidance. Taking a practical perspective, it covers a broad range of applications, including service robots. The text addresses the challenges involved in investigating methods by which PbD is used to provide robots with a generic and adaptive model of control. Drawing on findings from robot control, human-robot interaction, applied machine learning, artificial intelligence, and developmental and cognitive psychology, the book contains a large set of didactic and illustrative examples. Practical and comprehensive machine learning source codes are available on the book’s companion website: http://www.programming-by-demonstration.org

Autoren/Hrsg.

Calinon, Sylvain

Fachgebiete

• Technische Wissenschaften Elektronik | Nachrichtentechnik Elektronik Robotik

Inhaltsverzeichnis

ACKNOWLEDGMENTINTRODUCTIONContributionsOrganization of the bookReview of Robot Programming by Demonstration (PBD)Current state of the art in PbDSYSTEM ARCHITECTUREIllustration of the proposed probabilistic approachEncoding of motion in a Gaussian Mixture Model (GMM)Encoding of motion in Hidden Markov Model (HMM)Reproduction through Gaussian Mixture Regression (GMR)Reproduction by considering multiple constraintsLearning of model parametersReduction of dimensionality and latent space projectionModel selection and initializationRegularization of GMM parametersUse of prior information to speed up the learning processExtension to mixture models of varying density distributionsSummary of the chapterCOMPARISON AND OPTIMIZATION OF THE PARAMETERSOptimal reproduction of trajectories through HMM and GMM/GMROptimal latent space of motionOptimal selection of the number of GaussiansRobustness evaluation of the incremental learning processHANDLING OF CONSTRAINTS IN JOINT SPACE AND TASK SPACEInverse kinematicsHandling of task constraints in joint spaceexperiment with industrial robotHandling of task constraints in latent spaceexperiment with humanoid robotEXTENSION TO DYNAMICAL SYSTEM AND HANDLING OF PERTURBATIONSProposed dynamical systemInfluence of the dynamical system parametersExperimental setupExperimental resultsTRANSFERRING SKILLS THROUGH ACTIVE TEACHING METHODSExperimental setupExperimental resultsRoles of an active teaching scenarioUSING SOCIAL CUES TO SPEED UP THE LEARNING PROCESSExperimental setupExperimental resultsDISCUSSION, FUTURE WORK AND CONCLUSIONSAdvantages of the proposed approachFailures and limitations of the proposed approachFurther issuesFinal wordsREFERENCESINDEX