This post is also available in: heעברית (Hebrew)

Artificial Intelligence (AI) has been revolutionizing the video analytics field. One of the pioneers in the use of AI and deep learning technologies for a cloud-based video analytics service is Agent Video Intelligence (Agent Vi), a leading global provider of open architecture, video analytics solutions across various vertical markets, with over 4,500 installations in over 90 countries.

Agent Vi’s comprehensive video analytics offering includes software products for on-premise installations as well as cloud-based SaaS, with capabilities ranging from real-time video analysis and alerts to video search and business intelligence.

One of the main challenges facing video analytics has been object classification Deep Learning can be used to solve this problem by collecting many thousands of images from hundreds of surveillance cameras, which must first be labeled and classified, into a range of categories. To achieve the required accuracy rates, such a vast database must be collected and identified from actual surveillance footage. Once enough images are collected, a deep learning classifier algorithm can be trained and deployed as part of a video analytics solution, enabling it to practically eliminate most of the existing causes for false alarms.

Employing Agent Vi’s solutions allows users to improve security and safety in their facilities, enhance the efficiency of guards and surveillance systems operators, produce valuable business insights and better manage the traffic flow of people and vehicles.

The company’s solutions allow users to achieve optimal value from their video surveillance networks by automating video analysis to detect and alert for events of interest, expedite search in recorded video and extract statistical data from the footage captured by surveillance cameras.

One of the solutions offered by the company is innoVi for Smart Cities, a cloud-based innovative video analytics SaaS that transforms the city’s hundreds and even thousands of surveillance cameras into smart video devices and contributes to the city’s ability to improve security, safety and incident response city-wide.

Agent Vi’s solutions are fully integrated with a variety of cameras, encoders, video management systems and alarm automation software.

The company employs a unique, patented, distributed processing solution to perform initial processing of the video stream at the edge to reduce the amount of bandwidth required, and sends the results to the cloud service. A combination of CPUs and GPUs are then used to complete the processing and detect events. This unique distributed architecture overcomes the challenges associated with cloud-based video analytics including not being able to upload a high enough quality video stream to the cloud due to bandwidth limitations and not being able to deploy a GPU at the user sites due to cost considerations.

According to Itsik Kattan, Agent Video Intelligence CEO, one of the markets where the company believes video analytics can bring tremendous value is the Smart City market.

“The ability to use surveillance to detect security or terror threats, acts of violence, vandalism or other crimes, or even effectively monitor traffic – all rely on the ability to effectively, accurately and quickly be able to analyze the masses of video captured by the surveillance systems, while also being mindful of the citizens’ right to privacy. This is where video analytics becomes the core of a city’s ability to deliver on the whole Smart City concept.”

Video analytics allow city authorities and law enforcement to prevent incidents before they occur, or rapidly deploy an effective response while an incident is emerging.

The company believes video analytics should be affordable, practical to deploy, use and maintain over time, and adapt to the growing and changing needs of a city”  says Kattan. “That is why we have expanded our approach to deliver an entire infrastructure solution that covers everything from effective and secure access to the surveillance cameras, through video analysis, and to the accessibility of the analysis outputs to multiple relevant parties.” He sums up, “We are harnessing our field experience accumulated through over a decade of activity to deliver a real solution to real problems that would make it possible for any city to manage itself more effectively while delivering the highest levels of security and safety for the benefit of its citizens.”

Agent Video Intelligence participated in the Video Analytics 2017 Conference and Exhibition organized by iHLS.