Representation of Proprioceptive Inputs of the iCub Humanoid Robot Using Self-Organizing Maps
Reprezentace proprioceptivních vstupů humanoidního robota iCub pomocí samoorganizujích se map
Authors
Supervisors
Reviewers
Editors
Other contributors
Journal Title
Journal ISSN
Volume Title
Publisher
České vysoké učení technické v Praze
Czech Technical University in Prague
Czech Technical University in Prague
Date
Abstract
Cieľom tejto bakalárskej práce je vytvorenie biologicky inšpirovanej reprezentácie proprioceptívnych vstupov pre humanoidného robota iCub, študovaním a modelovaním procesov, ktoré sú zodpovedné za vývoj takýchto oblastí v primárnom somatosenzorickom kortexe mozgu u primátov. Nazbierané proprioceptívne dáta sú spracované pomocou rôznych typov populačného kódovania a následne pomocou Kohonenovej architektúry Samoorganizujúcich sa máp. Výsledky sú porovnané medzi sebou a tiež s proprioceptívnymi reprezentáciami u primátov. Dôraz je kladený najmä na zachovanie topológie vstupných dát, na pomerné zastúpenie vstupných dát vo výslednej mape a na celkovú reprezentáciu proprioceptívnych vstupov naprieč všetkými stupňami voľnosti.
The goal of this thesis is to create a biologically inspired representation of proprioceptive inputs for the humanoid robot iCub, studying and modeling processes that are responsible for the development of such areas in primary somatosensory cortex of primate brain. The proprioceptive data collected is preprocessed using different kinds of population coding and then fed into a Self-Organizing Map to create the representation. Results for different parameter settings are compared with proprioceptive representation in primate brain, focusing on topological organization of the map, proportional represenation of different values of inputs, and on the whole representation of proprioceptive inputs across all degrees of freedom.
The goal of this thesis is to create a biologically inspired representation of proprioceptive inputs for the humanoid robot iCub, studying and modeling processes that are responsible for the development of such areas in primary somatosensory cortex of primate brain. The proprioceptive data collected is preprocessed using different kinds of population coding and then fed into a Self-Organizing Map to create the representation. Results for different parameter settings are compared with proprioceptive representation in primate brain, focusing on topological organization of the map, proportional represenation of different values of inputs, and on the whole representation of proprioceptive inputs across all degrees of freedom.