Abstract:We propose here a consciousness framework for intelligent robot. This framework is defined as both synergistic modules of neural networks for outside inputs and reactions caused by consciousness movement. This framework reflects explicit connections between two parts of the system control unit: one global working memory and distributed modular cerebral networks relating to specific functions. Accordingly this framework is composed of three layers, physical mnemonic layer and abstract thinking layer, which cooperate together through a recognition layer to accomplish information storage and cognition. We consider two assumptions about this hybrid model: First, at any given time, many modular neural networks are running in parallel and process data from outside in an unconscious manner; Second, expressions are outputs of dynamic system driven by abstract thinking layer. Using this framework, various sorts of intelligent robot can be developed, leading to a general approach for computer-human interaction.

Neural Computing Consciousness Model For Intelligent Robot