Acoustic Pattern Recognition for Security Breaching Noise Detection

We propose the development of an acoustic recognition platform that can identify and locate specific sounds in high noise environments, and that can act as a foundation for non-line-of site detection and proactive alerts for other technologies such as cameras and UAVs. The proposed system is based on fundamental principles of neurobiological systems, i.e., it incorporates mathematical models and system architectures designed to mimic how the brain recognizes sounds from the outside world. The proposed system also will precisely localize the source of sounds of interest, and direct secondary response devices toward that source. The proposed system is an extension of a recently commercialized gunshot recognition and localization system, called SENTRI I, currently being sold in Chicago, Illinois, Tijuana, Mexico, and other US cities. The SENTRI I, however, is designed only to detect and recognize short-duration gunshots, and is not capable of analyzing and recognizing other security breaching noises such as RPGs and mortars, vehicle engines, human voices, etc. Re-design of this proven system are proposed to extend applicability to recognition and localization of a wider range of security-relevant noises, add new microphone technology to substantially extend the "listening range", and reduce the system footprint to a field-deployable size.BENEFITS: Development of adaptive, intelligent acoustic recognition/alert systems for security breaching noise detection, threat identification and perimeter protection using Safety Dynamics' core technology, the Dynamic Synapse Neural Network (DSNN). The DSNN technology has been optimized and is currently commercialized for gunshot recognition as the SENTRI (Smart Sensor Enabled Neural Threat Recognition and Identification), a system that recognizes gunshots sends range and bearing details to cameras which can then rotate and pan in on the source of the incident. The SENTRI is integrated with wireless communications, and a weather resistant box mounted on telephone or power poles for urban crime prevention. This STTR will extend Safety Dynamics' acoustic recognition technology to non-gunshot events critical to security breaching noises.