To make it the best user experience for you to work with the website, select your language.

Global

MEET
THE ALGORITHM

1ST PLACE
VERIFICATION ACCURACY
IDENTIFICATION SPEED
THE III BEST PERFORMANCE
ON THE DETECTION
OF PEDESTRIANS AND CYCLIST
THE 2ND BEST ACCURACY
PERFORMANCE ON DETECTION
ACTIVITY IN EXTENDED VIDEO

HOW FACIAL
IDENTIFICATION WORKS

How neural network works

The neural network is trained to identify unique facial features so that it can then find similar faces in the database. NtechLab’s face detection algorithm works with global facial databases, allowing for a split-second search.

Recognition accuracy FNMR=0.008 @FMR<10-6
250 millionimages under 0.2 seconds
500 millionimages under 0.3 seconds
1 billionimages under 0.5 seconds
Recognition speed<1 second
icon

When working with facial features it is impossible to restore the original image of the face, which ensures compliance with the rules of personal data protection.

icon

Working with facial features requires up to 100 times less computing power compared to face image processing.

RECOGNITION MECHANISM

01Detecting the face and silhouette in the image
02Visual distortion correction
03Retrieval of facial features
04Verification or identification of a face

The algorithm analyzes
an individual video frame

A video consists of frames. Each frame is nothing more than a collection of pixels.
Ресурс 5

Pixel color
determination

Each pixel has a unique color code. The color code in the RGB palette is represented as three numerical values.
icon

The neural network receives a matrix of RGB values of pixels as input.

33 333 333

The algorithm detects faces

icon

The algorithm determines where faces are on the image.

NtechLab’s face detection algorithm is capable of detecting an unlimited number of faces in a frame, making it an ideal solution for security in crowded areas.

The speed of the detector does not depend on the number of faces in the frame.

faces-border faces-border-mb-fullEn