DART

DART combines cutting-edge audio processing, data augmentation, and machine learning techniques to deliver state-of-the-art ML models for gunshot audio detection, classification, and localization. DART models are trained on the largest-reported gunshot audio dataset in the world using intricate augmentation techniques not previously reported by other researchers.

Person holding a tablet displaying a digital shooting range app with a map background and a list of ammunition types with percentages.
Close-up of a laptop mouse pad with blue illumination in a dark setting.
A drone flying in front of a landscape with trees and shrubs, captured from below with a tilt-shift effect.

Aircraft Classification Enabled by Machine Learning, ACE-ML

ACE-ML is a multipart project delivering state-of-the-art capabilities for helicopter and UAS detection, classification, and localization. Using de novo datasets and intricate augmentation techniques, ACE-ML allows for the creation of a low-cost, distributed sensor network to provide early warning systems and increased survivability to warfighters on 21st century battlefields.