Accession Number:

AD1105068

Title:

Integration of a Computational Auditory Scene Analysis Software Architecture with the ARCADIA Cognitive Model: Software Design, Usage, Implementation, Demonstration

Descriptive Note:

[Technical Report, Technical Report]

Corporate Author:

Naval Submarine Medical Research Laboratory

Personal Author(s):

• Gever, David
• Daley, Matthew
• Babina, Matthew
• Snihurowych, Bohdan
• Bonacci, Lia
• Bolkhovsky, Jeffrey

Report Date:

2020-07-28

Pagination or Media Count:

70

Abstract:

The Naval Submarine Medical Research Laboratory NSMRL has developed an auditory model, which is a software application that attempts to emulate human hearing. It can feed its results to cognitive models that attempt to mimic the behavior of humans and are able to take advantage of auditory information as if it came from a humans ears. The software application addresses the cocktail party problem, which occurs when humans are able to follow a particular conversation from within a diverse background of noise. The auditory model accomplishes this by implementing a series of signal processing steps that are inspired by human sensory processing, using Matlab and Clojure programming languages.

Descriptors:

• computer programming
• ear
• acoustic signals
• supervised machine learning
• software development
• computer languages
• software design
• machine learning
• information science
• application software
• cognition
• computational science
• computer programs
• computers
• acoustic properties
• computer science
• data mining
• feature extraction
• pattern recognition
• unmanned underwater vehicles

Subject Categories:

• Acoustics
• Anatomy and Physiology
• Computer Programming and Software

Distribution Statement:

[A, Approved For Public Release]