BriefCam announced that it has won the Defence and Security Accelerator (DASA) Hackathon. The event, sponsored by the UK government, was designed to simulate real-world incident response and was held on November 26-27, 2018 in London.
BriefCam earned the hackathon victory by using its next generation platform to successfully identify a real-world subject of interest in a terror simulation. BriefCam searched hundreds of hours of video from dozens of cameras and files, looking for a subject, based on limited known details. BriefCam was able to leverage low resolution, low frame-rate video files as well as simulated, high resolution, live streams to verify several images of the suspect. The subject of interest was accurately identified in less than two hours. The breadth of multimedia formats that are represented in real world investigations, particularly proprietary CCTV, provided a significant learning curve for everyone involved around the need for near real-time ingestion.
“Our mission is to help law enforcement agencies solve crimes faster and support municipalities in creating safer cities, so we are honoured to have won this hackathon, and demonstrate how we help accelerate the time-to-target for investigations,” said Trevor Matz, CEO of BriefCam. “By leveraging the advanced deep learning and artificial intelligence capabilities of our comprehensive video analytics platform, law enforcement agencies globally are protecting what matters most.”
All hackathon participants were asked to process large amounts of real-time, multimedia data, and rapidly identify key information such as people, places, and events that could assist an investigations team. The hackathon evaluated how subjects of interest relate to each other, the event timeline and narrative, and near-term predictions. Application of cutting-edge artificial intelligence (AI) techniques, including machine vision, was also considered.
BriefCam’s next generation video content analytics platform includes highly accurate face recognition seamlessly integrated with granular object classification, multi-camera search, quantitative analytics, and real-time alerting capabilities. Leveraging state-of-the-art Deep Neural Networks trained in diverse data sets for industry leading accuracy in face matching and object classification, BriefCam v5.3’s capabilities help streamline post-event investigations, support rapid response to situational changes, and derive richer operational intelligence for data-driven decision making.