Invited Tutorial: Metaheuristic Optimization With Applications To Computational Vision For Humanoid Robots

Nature-inspired computing studies the application of biology concepts to solve demanding problems by assuming that natural world experiences may hold some answers to real-life technical challenges. Engineering is being challenged everyday by more complex, large, illstructured and distributed systems yielding a renovated interest on the subject. However, nature is providing simple structures and organizations which are capable of dealing with most complex systems and tasks. Metaheuristic optimization techniques aims to mimic collective intelligence from several nature-inspired approaches in order to propose a solution for several problems exhibiting a complex behavioral pattern. The overall approach follows the idea that a system is composed of decentralized individuals that may effectively interact to other elements according to their localized knowledge, i. e. their individual interaction. Special kinds of artificial collective-individuals are the elements created by analogy with bees, charged particles or the human immunology system.