FINDFACE
MULTI

MULTI-OBJECT VIDEO ANALYTICS PLATFORM

FINDFACE MULTI

NtechLab platform processes video and recognizes human faces, bodies and actions, as well as cars and plate numbers. AI-powered technology enables record breaking accuracy and high speed of recognition. The multi-object and analytical capabilities of FindFace Multi unlock new scenarios for responding challenges of public sector and business.

Multi-Object Video Analytics

FindFace Multi quickly and accurately recognizes faces, human bodies, cars, and license plate numbers in a live video stream or in a video archive.

Searching for faces, bodies, and vehicles in a database or in an archive is available both by a photo sample and by specific features: for example, by age, clothes color, or vehicle model.

ACCURATE AND FAST ALGORITHM

NtechLab developers are constantly improving recognition algorithms, increasing their performance and accuracy*. With FindFace Multi it takes less than a second to detect a face in a video stream, recognize it, and search for a match in a database with billions of images. Algorithms reliably recognize the appearance of the same person in different images and videos, even under challenging conditions: for example after an appearance-change, at a long distance to an object, or in a video of low quality.

*FindFace Multi is based on kiwi—the fastest and the most accurate neural network in the world (According to NIST testing in May 2021)

Face mask

Match: 87%

Facial hair

Match: 89%

Turned or tilted head

Match: 87%

Headgear

Match: 92%

Glasses / sunglasses

Match: 85%

Aging

Match: 85%

LIVENESS

FindFace Multi has an integrated anti-spoofing system that works on any camera and distinguishes a live person from an image on paper or screen. The neural network analyzes several frames at once capturing changes on the face, thus ensuring a real person is present in front of the camera.

John Smith

Fake

FindFace Multi Liveness functionality is connected as a plug-in with API, so if anti-spoofing is not included the total cost is reduced.

FindFace Multi platform itself uses Liveness for user authorization by face.

LESS RESOURCES

FACE MASK DETECTION

Mask

FindFace Multi detects face mask presence and proper wearing. Additionally, it features high detection accuracy for faces in masks.

The system recognizes the presence of a face mask and detects three states:

No mask

Proper wearing

Improper wearing

SEARCH BY ATTRIBUTES

A search for a face, a body or a car by attribute filters takes a split second. High search performance unlocks efficient security scenarios: find the person or a car on videos by known features, trace the route, and identify relations in a seconds.

FACE ATTRIBUTES

GENDER

BEARD

MASK

AGE

GLASSES

EMOTIONS

BODY ATTRIBUTES

CLOTHES TYPE UPPER

CLOTHES COLOR UPPER

HEADWEAR

CLOTHES TYPE LOWER

CLOTHES COLOR LOWER

VEHICLE ATTRIBUTES

BODY TYPE

MAKE

BODY COLOR

LICENSE PLATE ATTRIBUTES

COUNTRY

REGION/STATE

SEARCH BY A PHOTO SAMPLE

The platform allows you to find images similar to a given photo sample in a video stream or a database. Such a search (re-identification) is available in FindFace Multi for faces, bodies, and cars.

With re-identification one can search video for people and cars for which there is no information available other than an image.

Cards

Cards are a convenient tool for organizing and storing information about objects and events recognized by the video analytics system. Cards support connections between objects, search, and more.

FindFace Multi automatically creates cards for every newly recognized unique object (person or vehicle), and then accumulates in a card all the new images associated with the ‘object’ from all connected cameras.

FACE AND BODY COUNT

In FindFace Multi, you can create a face / body counter with one or more connected cameras and find out the exact number of people passing through the coverage area. Even with multiple cameras, the program will not count the same person twice, making it possible to count people without errors, even in large spaces and in rooms of complex configuration.

The face and body count function can be used to monitor waiting areas to prevent queues and ensure people’s safety.

VISITOR ANALYTICS

The ability to analyze data and collect statistics on gender, age, emotions, and the number of unique and returning visitors allows you to improve customer service and customize product offerings to meet clients’ needs.

Accurate quantitative data collection about event visitors helps strengthen security and improve event organization and management.

RELATION ANALYSIS

The platform will discover the circle of contacts of an individual by recognizing faces in the same frame from a surveillance camera.

The analysis of human connections helps to search missing people, conduct digital investigations, or trace the path of infection transmission.

Video Recorder

Functionality for recording, storage, and playback of video data from cameras is embedded into FindFace Multi UI. From an event you can open a video playback window for the corresponding recording time.

With video recorder, customers can use FindFace Multi as a unified platform for working with video surveillance data within a single UI.

PRIVACY PROTECTION

FindFace Multi does not store information about people who are not in the database. You can automatically blur random faces in the frame.

Optional personal data protection protects visitors’ privacy without impact on video analytics productivity.

GENERATING REPORTS

FindFace Multi allows to generate any report with the click of a button. Go to the tab with the data you are interested in and request a report in .xlsx format. Thanks to the built-in manager, all your reports will always be at your fingertips.

Reports allow to use third-party BI software to analyze and visualize data without access to FindFace Multi installation.

SEAMLESS MIGRATION BETWEEN FINDFACE PLATFORM GENERATIONS

We offer a tool for a convenient upgrade to a new generation of the product.

FindFace
new versions

FindFace
early versions

When updating, all user settings and data are migrated to the new version of FindFace Multi.

The updated system is immediately ready for use.

FINDFACE MULTI API

Convenient and accessible API with detailed documentation and webhook support.

Easy and quick integration of FindFace Security API face recognition module into any third-party services

FindFace Multi API allows you to quickly integrate face, body, and car recognition modules into any third-party devices and services, such as access control systems, cash registers, monitoring systems, and many others.

EASY ACCESS RIGHTS MANAGEMENT

Efficient and convenient management of access rights to camera groups, watchlists, and events.

ADMINISTRATOR

SUPERVISOR

WATCHERS

Maintaining a structure with complex hierarchies is effortless. FindFace Multi ACL separates access rights to all functional parts of the system.

Fine-tune notifications on events: receive only those you need. For example, security staff will not receive alerts upon a VIP client visit.

Audit logs allow you to view the history of FindFace Multi users’ actions.

RANGE OF APPLICATION

Solutions that make the world a better place.

Crime prevention, search for offenders and missing people.

Access management, zoning, and vehicle control.

Reduction of financial losses.

Relevant offer for each of the clients.

Faster customer service with face and vehicle recognition.

Instant recognition of regular customers and identification of unwanted visitors.

PARTNERSHIP

PARTNER WITH NTECHLAB TO BECOME A PART OF THE FUTURE

Become an NtechLab provider, integrator or agent. Grow and earn with us.

We offer various options for collaboration in all areas of business. Submit an application and we will select the most attractive offer for you.

We offer partnership options for various types of partners. Send a request, and we will create an offer with the most attractive price and timing for implementing a system for multi-object video analytics.