Modeling Two-Channel Speech Processing With the EPIC Cognitive Architecture.

Area of Science:

• Cognitive Science
• Human-Computer Interaction
• Psychoacoustics

Background:

• Cognitive architectures model psychological mechanisms for predicting human performance in system designs.
• Modeling multitalker speech tasks presents a novel challenge for existing human performance models.

Purpose of the Study:

• To present a novel cognitive model for predicting human performance in a two-talker speech task.
• To incorporate psychoacoustic concepts into a computational model for human performance.

Main Methods:

• Developed a computational model based on cognitive architecture principles.
• Integrated concepts of auditory masking and stream formation from psychoacoustics.
• Applied the model to predict performance in a two-talker listening scenario.

Main Results:

• The model successfully incorporates psychoacoustic effects relevant to multitalker speech perception.
• Demonstrates a framework for predicting human performance in complex auditory environments.
• Provides a foundation for more accurate human-machine system design involving speech communication.

Conclusions:

• The proposed model offers a novel approach to understanding and predicting human performance in multitalker speech tasks.
• Integrating psychoacoustics enhances the fidelity of cognitive models for human-machine interaction.
• This work advances the application of cognitive architectures in designing systems that process complex auditory information.