JAIST Repository >
School of Information Science >
Conference Papers >
Conference Papers >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10119/14763

Title: A Memory-based Robot Architecture based on Contextual Information
Authors: Pratama, Ferdian
Mastrogiovanni, Fulvio
Chong, Nak Young
Keywords: memory-based architecture
contextual information
human-robot interface
Issue Date: 2014-12-10
Publisher: Italian Association for Artificial Intelligence
Magazine name: The thirteenth Symposium of the Italian Association for Artificial Intelligence
Abstract: In this paper, we present a preliminary conceptual design for a robot long-term memory architecture based on the notion of context. Contextual information is used to organize the data flow between Working Memory (including Perceptual Memory) and Long-Term Memory components. We discuss the major influence of the notion of context within Episodic Memory on Semantic and Procedural Memory, respectively. We address how the occurrence of specific object-related events in time impacts on the semantics associated with the representation of those events and the subsequent use of those objects through robot actions. The general architecture design and its implementation are described. A preliminary validation both in simulation and in a real-world environment is discussed. Current work includes an implementation on a Baxter dualarm manipulator.
Rights: Copyright (C) 2014 Authors. Ferdian Pratama, Fulvio Mastrogiovanni, Nak Young Chong, The thirteenth Symposium of the Italian Association for Artificial Intelligence, 2014.
URI: http://hdl.handle.net/10119/14763
Material Type: author
Appears in Collections:b11-1. 会議発表論文・発表資料 (Conference Papers)

Files in This Item:

File Description SizeFormat
22204.pdf467KbAdobe PDFView/Open

All items in DSpace are protected by copyright, with all rights reserved.


Contact : Library Information Section, Japan Advanced Institute of Science and Technology