Towards Parallel Selective Attention Using Psychophysiological States as the Basis for Functional Cognition.

Area of Science:

• Cognitive Science
• Artificial Intelligence
• Human-Computer Interaction

Background:

• Human-machine societies require believable agent behaviors.
• Cognitive processes like selective attention are crucial for efficient resource management in agents.
• Developing artificial general intelligence (AGI) and machine consciousness (MC) necessitates advanced cognitive architectures.

Purpose of the Study:

• To develop a Nature-inspired Humanoid Cognitive Computing Platform for Self-aware and Conscious Agents (NiHA).
• To investigate the impact of top-down and bottom-up attention mechanisms on service agents.
• To evaluate the behavioral influence of psychophysical states on attention within an agent.

Main Methods:

• Development of the NiHA cognitive architectural framework.
• Implementation of top-down and bottom-up attention mechanisms.
• Evaluation of psychophysical states' impact on parallel selective attention and contextual responses.

Main Results:

• The NiHA agent demonstrated high accuracy (approx. 90%) in attention generation.
• The agent achieved 89% accuracy in attention within social interactions, considering psychophysical states.
• Integrating emotions into the attention process improved contextual-based responses.

Conclusions:

• NiHA provides a framework for advancing machine consciousness and AGI.
• Selective attention mechanisms, influenced by psychophysical states and emotions, are vital for believable agent behavior.
• The study highlights the importance of integrating cognitive and emotional processes for sophisticated AI agents.