NtechLab integrated computer vision into the 112-Emergency Aid application of the Novosibirsk Region

2021: Introduction of computer vision for processing photos of the 112-Emergency Assistance application

The use of technologies based on neural networks has significantly reduced the load on operators of Service-112Novosbir region. This was made possible thanks to the cooperation of the Ministry of Digital Development and Communications of the region with one of the representatives in the field of solutions for intelligent video analytics - the technological partner of Rostec State Corporation, the IT company NtechLab. As a result, the load on the system operators decreased by 10 times. This was announced on March 17, 2021 by NtechLab.

Since 2020 Novosibirsk mobile application , the section "" has been introduced in the region in 112-Emergency Assistance. COVID-19 The application is positioned, first of all, as a means of self-monitoring and monitoring the state of health of Novosibirsk people who are treated for coronavirus outpatients. They can enter information their health information in the application, receive consultation from doctors, and in case of a sharp deterioration in health, call an ambulance medical. Another important part of the functionality is the confirmation of compliance with the self-isolation regime - a selfie attached to the health questionnaire is compared with a reference photo in the user's questionnaire.

At the first stage, operators made a manual comparison of photos - they had to view thousands of photos a day daily, which led to verification delays. To automate this process and unload operators, the Ministry of Digital Development and Communications of the Novosibirsk Region turned to the company NtechLab with a proposal to integrate the functions of comparing selfies of users and obtaining a degree of similarity of photos. We are talking about the use of technologies of artificial neural networks (computer vision) in the mobile application "112-Emergency Assistance," which is one of the modules of the System-112 of the Novosibirsk region.

The use of computer vision was able to facilitate the process of staying on self-isolation for thousands of Novosibirsk residents, reducing the verification time to a minimum. No less important is the level of protection of personal data. Thanks to our cooperation with leading IT companies, we significantly improve the quality of digital services provided to residents of the Novosibirsk region, "said Anatoly Dyubanov, Minister of Digital Development and Communications of the region.

The total number of COVID-19 patients who sent selfies while on self-isolation in 2020-2021 exceeded 13,000 people. During this period, they sent operators more than 244,000 selfies - operators needed to process an average of 300-400 photos per day.

Video analytics technologies using artificial neural network technologies NtechLab made it possible to completely automate the process. The deep technological integration of the company's algorithms with the software solution was carried out in a short time - less than a month. This made it possible to unload the work of operators and significantly accelerate the process of identifying people in photographs. After the introduction of computer vision, it began to pass automatically, in less than one second, and as of March 2021, operators manually look at no more than 30-40 photos per day. A person connects only when the photo is not clear enough and additional confirmation by the human eye is required.

Anatoly Dyubanov stressed that within the framework of mutual cooperation, it is planned to further scale computer vision in other areas where identity confirmation is required through sending photos.

2020: Plans to launch a facial recognition project

In two cities of the Novosibirsk region will launch a pilot project on the introduction of a face recognition system. This became known on July 6, 2020.

The experiment will be conducted in the cities of Toguchin and Karasuk.

At the beginning of July 2020, the Netris CCTV video surveillance system platform was deployed in the Data Processing Center of the Government of the Novosibirsk Region. It is part of the "Regional Monitoring System for the Components of Ensuring Public Safety, Municipal Infrastructure and Environmental Safety in the Novosibirsk Region."

With the help of the service, data are collected and processed from various automated systems for ensuring the safety of citizens for forecasting, monitoring threats of a natural, man-made, social and environmental nature in the region.

The Ministry of Digital Development of the region in test mode carried out work on integration with the video surveillance system of facial recognition algorithms NtechLab, and work is also underway on the integration of VisionLabs algorithms and the Center for Speech Technologies LLC. Licenses have been purchased for up to 1000 CCTV cameras, as well as Netris intelligent video processing module with the ability to connect up to 2000 devices.

At the beginning of July 2020, more than 200 cameras were connected to the video surveillance system. They are installed at objects with a mass stay of people. In the future, all educational organizations of the region, as well as state budgetary institutions of health care, plan to connect to the system.

For the implementation of the pilot project, facial recognition algorithms were provided by companies free of charge until the end of 2020. As a result of the work, a decision will be made to finance the acquisition of licenses from development companies^[1]

Notes