Solana Networks Conducts State of the Art Research into Deep Learning Applications for Defence R&D Canada
Solana Networks completed a project for Defence R&D Canada (DRDC) conducting advanced research into Deep Learning applications in support of Threat Anticipation and Situational Awareness.
The project focused on included a detailed study of state of the art predictive analytics methods and research along with development of two software tools applying AI for situational awareness in the maritime domain. Investigation was carried out into Pattern-of-Life (POL) detection. The AI tools included (i) a Labeler tool used to facilitate rapid labeling of maritime vessel trajectory data for use as training and ground-truth in AI models (ii) A Modeler tool which enabled building & evaluation of AI models to detect and classify maritime vessel trajectories.
Accurately labeled datasets are the cornerstone for training of AI algorithms. The Labeler tool developed in the project greatly expedites the work of AI practitioners to rapidly build labeled datasets for AI model building in different domains including image classification, object detection, cyber security and network traffic classification among others. The Modeler tool developed in the project utilizes a novel method to address hard-to-classify trajectories for maritime vessels.