Using AI to Augment NG9-1-1 Operational Efficiency by Detecting Non-emergency and Anomalous Video Content

SecureLogix is working on several projects with the DHS to secure NG9-1-1. The focus has been on threats such as Telephony Denial of Service (TDoS), robocalls, and SWATing. This architecture is based upon our PolicyGuru and Orchestra One solutions. These solutions include an architecture that can easily can be deployed in NG9-1-1 and expanded to provide access to the multimedia data, including video. As part of a separate project, this architecture is being enhanced with an AI/ML infrastructure, which can be easily expanded with new video analysis models.SecureLogix proposes to build a prototype video analysis system, which will reduce the burden and workload of the telecommunicator. Video allows sending very large and redundant content, which will result in high cognitive load on telecommunicators. We propose to develop a system that will significantly reduce the cognitive load and allow for efficient response. The cognitive load will be reduced by processing the video and transforming it into a set of analyzed artifacts that are easier to consume. Video segmentation, summarization and object detection will be applied. DeepFake and spam detectors will also be used. The telecommunicator will have access to a set of representative frames and a short summary video, which will preserve the chronological order of the important scenes in the original video. This will save precious time for the telecommunicator and dispatched resources in the field.The results of the video analysis will be integrated with the SecureLogix solution and summarized globally using the Intrado ECaTS dashboard.