Visuospatial Cognition and Episodic Memory in Synthetic Characters
In studies of human memory, the cognitive structures and processes involved in binding visual, spatial, and episodic information are essential in visuospatial cognition and indissociable from mechanisms such as visuospatial sketchpad, episodic buffer, and the episodic memory. Such visuospatial and episodic-related features are essential to deal with qualitative spatial representations. The aim of this work is to develop a cognitive architecture inspired by these structures and processes in order to provide synthetic characters (as in the case of computer games or social simulations) with cognitive skills and abilities in which they might be able to create and use descriptive-map-based representations to deal with self-location, mapping, and visuospatial-related communication problems. Our plan is to design and develop a codelet-based cognitive architecture, and conduct experiments in order to evaluate synthetic characters' performance. Our goal is to propose a robust and scalable cognitive architecture that supports binding multimodal information and provides synthetic characters with human-like descriptive-map-based self-location, mapping and communication skills and abilities.
Goals
The aim of this work can be paraphrased as follows: how to endow synthetic characters with visuospatial cognition and episodic memory to be able to produce and use human-like descriptive maps. The research plan to reach a possible bio-inspired response to the question above has two strands: (1) design and develop a codelet-based cognitive architecture, and (2) conduct experiments evaluating the performance of synthetic characters governed by such cognitive architecture. More precisely, the first strand focuses on (1) Conceptualizing and developing computational components inspired on the structures and processes related to the binding of visual, spatial and episodic information. (2) Implementing a multimodal information flow structure. (3) Developing and implementing the cognitive architecture based on those computational components and the information flow structure. Finally, the second strand focuses on the cognitive architecture evaluation. For the implementation of the cognitive architecture as mentioned in (3) we will use the CST - Cognitive System Toolkit (CST) - which is founded on concepts of Global Workspace Theory to implement the codelets, coalitions, structures and processes.
Authors and Contributors
. . . .
Contact
. . . .