Enhanced Identity Object Model (EIOM) for High Value Target Identification

ROC proposes to develop the Enhanced Identity Object Model (EIOM) from the successful Phase 1 feasibility to a Phase II software prototype for Army T&E that incorporates both face and non-face feature imagery into an enhanced biometric template that can effectively track an individual and re-identify that person in video from multiple cameras with non-overlapping coverages. ROC will leverage its top-tier face recognition algorithm to assess the saliency of including non-face features to improve tracking performance. During Phase 2, ROC proposes to: 1. Develop algorithmic efficiency and accuracy of EIOM using ROC AI/ML methods proven for high accuracy and low error rates in the National Institute of Standards and Technology (NIST), Face Recognition Technology Evaluation (FRTE) and Face Analysis Technology Evaluation (FATE). 2. Rigorously test and evaluate the EIOM against large volumes of streaming video. 3. Package the EIOM in a software SDK for end-user test and evaluation. ROCÆs commercial face recognition SDK has been proven for ease of integration in many hundreds of installations across government and commercial customers. Headquartered in Denver, CO and made in the USA, Rank One Computing is an industry-leading developer of AI/ML-based video analytics including face recognition, fingerprint matching, object detection, and optical character recognition algorithms.