Područje interesa Odjela za računalnu inteligenciju je teorija, oblikovanje, primjena i razvoj biološki i lingvistički motiviranih računalnih paradigmi s naglaskom na neuronske mreže, konekcijske sustave, genetičke algoritme, evolucijsko programiranje, neizrazite sustave i hibridne inteligentne sustave koji su temeljeni na ovim paradigmama.
Odjel za računalnu inteligenciju
Zavod za automatiku i računalno inženjerstvo, KoREMA i IEEE Computational Intelligence Society Chapter pozivaju vas na predavanje: "Multi-Joint Action - Towards Human-Robot Teams" koje će održati dr. Dražen Brščić s Technische Universität München, Njemačka.
Predavanje će biti održano u ponedjeljak 25. listopada 2010. godine s početkom u 14:00 sati u Sivoj vjećnici.
The abstract:
The research area of "multi joint action" is concerned with actions in which multiple cognitive systems take part. We are especially interested in multi joint actions that change the physical state of the world, where by cognitive systems we mean primarily humans and robots, with the final goal being the study of mixed and heterogeneous human-robot teams.
Due to the presence of multiple actors, and interdependence and mutual influence of their actions, it is necessary to take a different perspective on multi joint action compared to actions taken by a single cognitive system. The distinctive cognitive aspects of multi joint action can be defined as: 1) shared knowledge (perception and communication, 2) coordination (self- and team-awareness), and 3) joint execution (predictability and adaptation). The first aspect stands for the part of the knowledge about the joint action that is shared, i.e. existing on all involved cognitive systems. The channels for acquisition and sharing of this knowledge are perception, which is similar to the single system case, and direct sharing through communication, which is typical for joint action. In the planning and coordination of actions inside a team important roles are played by the awareness of own capabilities, as well as awareness of the team partners’ capabilities, actions, intentions, habits, etc. The last aspect defines the direct influences during joint execution: predictability is concerned both with the prediction of actions of the partner, as well as with the performing in a way that is predictable to the partner; adaptation includes the changes in own actions as immediate response to detected characteristics of the team or changes thereof.
This talk will give an overview of the work done on the topic of multi joint action as part of the German cluster of excellence CoTeSys - Cognition for Technical Systems. The presentation will cover the above mentioned three aspects of multi joint action and give details on the specific research topics and achieved results.
Kratki životopis predavača:
Dražen Brščić diplomirao 2000. te magistrirao 2004. godine na Fakultetu elektrotehnike i računarstva Sveučilišta u Zagrebu. Doktorirao je na Tokijskom sveučilištu u Japanu 2008. godine. Trenutno je uposlen kao post-doktorski istraživač na Technische Universität München, Njemačka. Istraživački interesi dr. Brščića uključuju mobilnu robotiku, distribuiranu percepciju, fuziju senzora, inteligentne prostore, te interakciju ljudi i robota.