Title :
Novelty gated episodic memory formation for robot exploration
Author :
Barakova, Emilia I. ; Lourens, Tino
Author_Institution :
Brain Sci. Inst., RIKEN, Saitama, Japan
Abstract :
This paper presents a method for novelty and familiarity detection, aiming at inferential use of episodic memories for modeling behavior in novel situations. The method is based on the simulation of the hippocampal function, especially on its aspects, that relate to the memory formation in spatial context as: (1) the sensory, perceptual, and behavioral correlates of the episodic memory formation, (2) its involvement in novelty/familiarity detection and inferential reuse of old memories, and (3) the natural way to relate the internal hippocampal and abstract spatial representations. The study differs substantially from the existing models, that relate the hippocampal function and robot exploration, since it focuses on flexible reuse of experienced episodes rather than on navigation. The model is built on the experimentally supported hypothesis of the novelty/familiarity discrimination function of the hippocampal area CA1.
Keywords :
brain models; cognitive systems; computerised navigation; inference mechanisms; mobile robots; abstract spatial representations; behavioral memory formation; episodic memory formation; familiarity detection; flexible reuse; hippocampal function; perceptual memory formation; robot detection; robot exploration; robotics; sensory memory formation; Artificial biological organs; Brain modeling; Context modeling; Encoding; Humans; Intelligent robots; Navigation; Organisms; Robot sensing systems; Service robots;
Conference_Titel :
Intelligent Systems, 2004. Proceedings. 2004 2nd International IEEE Conference
Print_ISBN :
0-7803-8278-1
DOI :
10.1109/IS.2004.1344648