RAIC-Next Generation AI Automation and ATR using Rapid Automatic Image Categorization

For this proposal, Synthetaic will adapt its early-stage Rapid Automated Imagery Categorization (RAIC) commercial product to enable rapid automatic target detection (ATD) and automatic target recognition (ATD/ATR) in support of Joint All Domain Command and Control (JADC2) initiatives. During the 12-month R/R&D period and 3-month documentation period, Synthetaic will work closely with the 110th Wing, Michigan Air National Guard, other DAF units, and an Army unit to adapt RAIC to demonstrate that any user can build AI models in minutes for immediate use in ATD/ATR activities. The proposed Phase II work will demonstrate that any DAF end user can build and deploy AI in minutes with no pre-labeled training data and then detect and identify targets hidden inside millions of images or hundreds of hours of MQ-9 Reaper video (in seconds). Once a model is built, any user with access to RAIC and data from the same SAR, FMV, or EO sensor can use the model to search new imagery.  The proposed R/R&D work includes adapting RAIC for use on SAR and FMV data for military applications and maturing RAIC's commercial cloud architecture and user interface to enable demonstrations and evaluation during DAF exercises (e.g., Northern Strike).