With the machine learning and AI stack getting more and more powerful, it enables the reduction of the domain complexity to be visualized as data points that could be crunched and analysed. The multidimensional video signals which we know as picture in motion, is today conferred as matrices for patterns and differentiators. Our data scientists and algorists are capable of building NN or Deep Learning modules that dive into the domain semantics and extract features on which intelligent analytics can be performed.
With intelligent systems empowered to analyse the video feeds the implication and impacts see no bound in our day to day life. For example in the entertainment industry, we could classify the amateur videos on the social network onto genres thus respecting the copyrights and parental control. Thus the copyright and licensing would become part of the video intelligence thereby self classifying and mutating the viewer rights and restrictions with respect to the content, when and wherever appropriate.
In the social security, when today we have human monitored CCTV, the added intelligence module looks out for malicious and suspicious behavior, thereby triggering a warning to the authorities who would probably look out to the matter much closer. Face recognition is one technique that arms the video analytics to pin down on look out suspects at transit locations and public gateways.
With the upcoming boom of the smart cities, the contribution of video intelligence is inevitable. We could implement decision modules that regulate and pre-plan the traffic queues and monitor the tolls and signals, all based on the information from the live feeds. The dash cam views of the connected vehicles could be harnessed to identify violations, suggest the cruising speed and recommend detours to avoid blocks and jams.With this technology we aim at not replacing or substituting the current authorities but empowering them to seal the leaks and human misses.