University of Twente Student Theses


Bringing the iCub to life with neural network activation

Roppelt, J. C. (2016) Bringing the iCub to life with neural network activation.

[img] PDF
Abstract:Robots are getting more and more sophisticated and involved in our daily lives. With growing complexity of tasks to fulfill, the importance of technical cognition is on the rise. The field of technical cognition does not only enhance robot behavior but can also help us to further our understanding of the human brain. The robot discussed in this thesis is the humanoid iCub. The iCub is developed to impersonate a toddler, both in physical and psychological terms. One of the most important characteristics of a human is his efficient and sustainable learning. There are numerous artificial neural networks that try to simulate the learning process of the brain. This thesis uses the neural network architecture of the motor-blackboard. The theory is discussed and implemented in a simulation, which is then integrated to work with the iCub as input and output device. It is shown that the motor-blackboard architecture provides a good basis for a sustainable learning program. Further research should be done on how values have to be set and how rhythms affect human movements.
Item Type:Essay (Master)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:77 psychology
Programme:Psychology MSc (66604)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page