Time KYC
TIME KYC Service
:
ABOUT THE SYSTEM
00
01
Time KYC:
Know Your Customer
Know Your Customer
TIME KYC Service is designed for automation of access control process and face tracking, to increase security by using face identification, to prevent security violations and to create a common information space for video content analysis
Turnkey solutionComplex solution with detailed user interface, mobile app, API for integration with existing systems and 1C integration module.
World's best algorythmsAccuracy is acknowledged by independent international tests carried out among vendors all over the world
TPR al FPR 10^-3 = 0.9915
TPR al FPR 10^-5 = 0.9909
High speedRecognition process takes up to 2 seconds from receiving a shot from the camera to sending a notification to the user interface.
Biometric template (descriptor) matching speed: 5 million matches/sec per CPU core
Fully protected dataReverse transformation of biometric template to image is not possible.
Photo images and personal data are not stored.
Flexible configurationEach service inside the platform works independently. The platform can be deployed on one or more servers with the capability of horizontal scaling.
Autonomous processingVideo data can be processed autonomously without transmitting big amounts of data through the Internet. The platform can be also installed on virtual machines in the company's cloud infrastructure.
:
SERVICES
00
02
We offer development cycle services: from business intelligence to development and support of the project post-launch.
:
CAPABILITIES
00
03
Security
and transparency
and transparency
What do you get with
Time KYC PLATFORM deployment in your company.
Time KYC PLATFORM deployment in your company.
- SecurityGuaranteed security level enhancement by incident response time reduction, increase of employee discipline.
- Access restrictionPrevention of access violations by unauthorized persons.
- Security cost reductionVideoarchive analysis cost reduced via automation of the process.
- Employee trackingPeople in the facility can be located at any time, information about movement tracks is constantly recorded.
- Anti-terrorism securityReduced terrorism threat, risk of facility access violation with fake identity and smartcards by using additional face identification.
Identification, geolocation and data collection
What capabilities offers the platform
Real-time videoReal-time processing of current video data and retrospective analysis of video files.
Employees trackingThe system locates people onsite after analyzing the information from the cameras using facial recognition technology in real time.
General visualizationVisualization of the facility territory with the number of persons in every controlled zone.
Employee's trackCheck employee's movement and itineraries for the desired period of time - colored tracks will be shown on the map.
ReportsThe system is capable of building various reports. It is possible to get analytical reports and statistics (for events, stored templates, user and object data)
Reports regarding time spent by an employee in different areas are available.
NotificationsStaff members receive notifications about important events, such as suspicious or unauthorized person, or VIP client appearance, on their PC or to the mobile app.
Work time trackingEmployees work time tracking. The system starts to pay off.
Access permissionsAll the actions of the system users can be logged. Access permissions can be differentiated.
Restricted areasPassive surveillance in the restricted areas that helps to detect unauthorized persons.
Smartcard-free accessAccess of the employees to the offices or restricted areas using a face - smartcard-free access control system.
:
BUSINESS SOLUTIONS
00
04
Time KYC Turnkey packages
for different types of business
for different types of business
The use of face recognition for bank office visitors and ATM's users will make the communication with the bank friendly and secure.
Clients will appreciate individual approach and employees will work more effectively.
Detect suspicious behaviour and identify persons from the blacklist
Detailed statistics of visitors and their behaviour
Detailed statistics of visitors and their behaviour
The most up-to-date security system based on face recognition algorithms
Integration with fire systems
Integration with fire systems
All passengers and visitors are being constantly checked in real time by comparing with the images from blacklist.
Effective targeted marketing and lost people search.
Effective targeted marketing and lost people search.
Simple contactless access control system based on face recognition algorithms
Workplace security, employee discipline and work time tracking
Workplace security, employee discipline and work time tracking
Big data analysis about movements and preferences of visitors.
Effective instruments that help to convert an occasional customer into a regular one.
Effective instruments that help to convert an occasional customer into a regular one.
Can't find your case?
We'll be happy to offer you a personal solution.
We'll be happy to offer you a personal solution.
:
HOW DOES THE PLATFORM WORK?
00
05
Face recognition
The platform solves the task to identify a person using a face image. For each image at the system's input a so called "descriptor" is calculated, which is used for comparison with other image descriptors. Face verification and identification are based on those comparison operations using one pair or multiple pairs of descriptors.
The process of face identification and matching with the biometric profile in the database.
At this stage faces are detected in the camera view area, face image parameters are estimated (light level, blurriness, head pose angles), liveness checks are performed. As a result of face detection and tracking a so called "best shot" is taken which is ready for biometric template (descriptor) extraction. Face detection can be performed in two modes: cooperative and non-cooperative. The cooperative mode means that a person interacts with the system to get a better face image and help the recognition process. In the non-cooperative mode a person does not interact with the system in any way and the recognition process is performed in the background.
At the biometric template extraction stage the face image is used as an input data for a pre-trained neural network algorithm which converts it to a compact vector of values using a set of convolution operations. Those values describe the image's features.
At the matching (comparison) stage the distance between two templates is estimated to get the information about the similarity of the faces. Less distance between the templates means higher the similarity.
:
ARCHITECTURE
00
06
Functional modules of the system
The system basically consists of these key subsystems: videodata processing module, web application and biometry module.
Installed IP camera is used to get face images of the users. The module sends detected face images to the system for matching with the existing database. The matching result and the information about the face image is saved for further use.
Biometry module is designed to solve these main tasks:
- descriptor extraction from the face image;
- descriptor matching 1:M.
- external API;
- descriptor extraction from the face images;
- descriptor matching;
- database management service.
The system consists of the following functional blocks:
Functional block "Analytics" is used to build statistics in table format, graphs and for data export.
Functional block "Search" is used to search a person using uploaded image or person's ID.
Functional block "Movement control" is used for notification of the users via GUI and mobile app and for person's track visualization on the facility map.
Functional block "Administration" is used for system user creation and provides the Lists section management.
- Lists;
- Analytics;
- Search;
- Monitoring and movement control;
- Administration.
Functional block "Analytics" is used to build statistics in table format, graphs and for data export.
Functional block "Search" is used to search a person using uploaded image or person's ID.
Functional block "Movement control" is used for notification of the users via GUI and mobile app and for person's track visualization on the facility map.
Functional block "Administration" is used for system user creation and provides the Lists section management.
:
CLIENTS AND PARTNERS
00
07
Implemented projects and technologies
Tochka BankTochka Bank. Implementation of the remote biometric identification system for 3,000 users of mobile app.
Pochta bankImplementation of the biometric identification system for loan application process control in 1,000 bank offices.
Ministry of agriculture and fishery of Astrakhan regionDeployment and support of accounting and finance systems for more than 10,000 users.
Ministry of Social Development and Labor of the Astrakhan regionDeployment and support of accounting and finance systems for more than 3,000 users.
Ministry of education of Astrakhan regionDeployment and support of accounting and finance systems for more than 8,000 users.
VisionLabsAlgorithms used by AstraTime for the platform of biometric monitoring are implemented in biometric identification systems of "Sberbank" and "Kaspi bank".
TevianAlgorithms used by AstraTime for the platform of biometric monitoring are implemented in the companies working in the telecommunications area.
AyonixFastest algorithm based on 3D face recognition.
3DiViAlgorithm of real-time pose tracking helps to detect suspicious behaviour.
:
ABOUT US
00
08
AstraTime LLC
The company was founded in 2019; the head office is located in Moscow.