Select

Real-time Internal and External Behaviour at Source Detection Using In-Car Video AI Sensors for Complex Event Detection

Technology Overview

This technology extracts information about linked activities from video feeds in near real time without human intervention. The models currently trained using video precognition Artificial Intelligence (AI) technology focused on identifying human-based activity but are equally adept at detecting other sequences of behaviour from sources other than people, simply by retraining the underlying Recurring Neural Networks (Deep Learning). This allows the system to be used for detecting complex behaviours such as: Anti-social behaviour (fighting in the street, threatening behaviour etc) Littering Criminal activity such as suspicious looking gatherings or behaviours (loitering / running etc) Traffic congestion including identifying whether it is simple congestion or the result of some sort of road incident or collision Monitoring of parking areas with identification of entry / exit for example arrival / departure of public transport,monitoringthe safety of passengers getting on / off, and any suspicious packages being left on platforms A mobile 'video precognition onboard' equipped camera (for example mandated to be deployed in taxis) would allow a wealth of valuable data generatedfor connected car development, such as being able to understand the relationship between driver distraction and the ability of the driver to resume control of the vehicle. The mobile solution would also make it far harder for criminals to avoid being observed by keeping out of sight of the fixed cameras Potential users include: Muncipal authorities Transport authorities Emergency services Automotive OEM & Non-OEMs Fleet companies Mobility Providers Insurance companies

Technology Features & Specifications

Video analytics transformation ismaking it possible to predict likely event outcomes. Surveillance videos can be monitored automatically in real time, triggering alerts to be sent to response teams such as emergency services as required. This allows crowd managers and emergency services to target resources effectively. Current video analytic solutions are limited by their capability to capture relevant information from video streams in real-time. Human operators can only monitor a limited number of screens and their reliability decreases dramatically after 20 minutes continuous monitoring. The performance of automatic video analytic solutions is also limited, with high rates of false alarms and missed events. By approaching video analytics in an entirely new way, the technology has succeeded in realizing the concept of video precognition. This system is able to predict when a fight is about to break out in a crowd, or identify a car drivingtowards people. The system scales easily to process thousands of hours of videos and trained to recognise a large array of different behaviours. Expertise in big data solutions enables the delivery of an innovative, robust and computationally efficient video precognition system that can interface easily with third-party systems. The specification includes: AI GPU Supercomputer Data processing modules Automated Data Analysis Automated Control & Optimisation Systems Connection to in-car technology enablers Connection to infrastructure Technology comprises: Video cameras Autonomous car sensors e.g. LIDAR MIMO interface Control systems Algorithms AI-GPU Chipset

Potential Applications

A mobile 'video precognition onboard' equipped camera (for example mandated to be deployed in taxis) would allow a wealth of valuable data generatedfor connected car development, such as being able to understand the relationship between driver distraction and the ability of the driver to resume control of the vehicle. Anti-social behaviour (fighting in the street, threatening behaviour etc) Litter monitoring and alert Criminal activity such as suspicious-looking gatherings or behaviours (loitering / running etc.) Traffic congestion, such as identifying whether it is simple congestion or the result of some sort of road incident or collision Monitoring of parking areas with identification of entry / exit for example Monitoring of arrival / departure of public transport,monitoringthe safety of passengers getting on / off, and any suspicious packages being left on platforms The mobile solution would also make it far harder for criminals to avoid being observed by keeping out of sight of the fixed cameras

Customer Benefits

IoT applications usually offer more value when they incorporate video analytics, since the technology allows them to consider a wider range of inputs and make more sophisticated decisions. For instance, some typical IoT applications use beacons that transmit location data each time they connect with a consumer smartphone in a store. Using deep learning to analyse live and batch video data to predict events before they occur, the technology provides accurate and real-time insights for different applications including emergency services, e-commerce, advertising and surveillance, as well as connected and autonomous vehicles. For example, focusing on crowd dynamics, the technology provides insights into numbers, clusters, distributions, movements and behaviours.Concurrently, a big data analytics back-end provides the real-time data and insights needed to inform operational decision-making and risk management. Customer benefits include: Insights from video surpassing human limitations Rapidly recognizes targeted actions in live and historical video streams Picks up things which the human eye misses Rapidly decreases response times Informs what is happening, enabling immediate response to critical scenarios saving precious time Providescritical information instantly: Automatically alerts decision makers and people on the ground Emergency services can react to situations before its too late

Make an Enquiry