SpectralAi Nexus employs Deep Learning Technology to convert your Scribe target annotations into a trained Convolutional Neural Network (CNN) model. 

These models are then used for automatic target recognition (ATR) processing of survey data. The models generated are proprietary to the organization and data. 

As targets are annotated from additional survey data, Nexus will continually improve the models by re-training in Nexus. Best of all, Nexus requires no internet connection! Nexus runs on local NVIDIA GPU-based hardware.

Nexus provides comprehensive target management enabling multiple projects to be merged into a single solution. The targets from each project are loaded into the Target Catalog and Target Editor. 

The Target Catalog filters and refines the final target dataset used to train the CNN model. Nexus also enables further target editing prior to training and will display a target in context in a high-resolution waterfall display.