Владельцам экранов в ритейле
Рекламодателям
О компании
Укажите детали обращения
Заполните поля, и наш эксперт свяжется с вами для проведения демо
Нажимая на кнопку, вы соглашаетесь с условиями о персональных данных
© 2016 Your company
Screen Owners
Advertisers
About
For Advertisers
How it works?
For DSP
Use Cases
About Us
© 2016 Your company
Now you can provide them with accurate and confirmed numbers in real time.
Advertisers & marketers require real numbers
Audience
Engagement
Conversion
Frequency
Reports

Get real-time impersonal
data in physical locations

Audience behaviour
Emotion tracking
Attention on Ads
Customer Journey
Mask detection

Identify face mask usage in real-time to alert your customers about safety rules

Put on a mask
please
AI algorithms specially designed and developed for market fit, restrictions and to be relevant in real cases with easy to set up.
Out of the box
Best recognition algorithms
One of the best algorithms in the world confirmed by multiple global independent tests. Extremely robust and accurate. 98%+ accuracy of gender recognition, 95%+ accuracy of age group recognition.
Full cycle
Algorithms allow working with the audience. Detection, Tracking, Recognition and Identification. All necessary technical solutions in one box work together.
Self-sufficient
No need custom integrations with third-party tools. Video analytics work out-of-the-box simultaneously with content playback without extra devices. Just plug-in camera.
Data sets
Own data sets and a neural network that used machine learning to be trained with millions of images by real cases.
No special hardware
No additional preparations to start. Just use simple USB cameras to work with your audience.
Big Data
Dozens of advertising and marketing reports are available by provided recognized audience.
Best recognition algorithms
One of the best algorithms in the world confirmed by multiple global independent tests. Extremely robust and accurate. 98%+ accuracy of gender recognition, 95%+ accuracy of age group recognition.
Full cycle
Algorithms allow working with the audience. Detection, Tracking, Recognition and Identification. All necessary technical solutions in one box work together.
Self-sufficient
No need custom integrations with third-party tools. Video analytics work out-of-the-box simultaneously with content playback without extra devices. Just plug-in camera.
Data sets
Own data sets and a neural network that used machine learning to be trained with millions of images by real cases.
No special hardware
No additional preparations to start. Just use simple USB cameras to work with your audience.
Big Data
Dozens of advertising and marketing reports are available by provided recognized audience.

Our computer vision algorithms suited to work in real cases locally even without network using basic cameras and recognize 12+ different attributes.

What our AI see?
1
Only real faces
Algorithms trained against faces spoofing (faces on POS materials, video walls, other screens, T-Shirts and so on).
2
Mask detection
Faces in masks detection support. Do not "loose" your audience even in COVID time.
3
Staff exclusion
Separate staff from statistics data. Staff will be marked to exclude it from the main analytics data. No selling ads content to them if you don't want it.
4
Hair detection
Facial hair and hairstyle and hair colour recognition.
5
Accessories detection
Glasses and headwear recognition.
6
Group recognition
No faces count limits. The algorithm can recognize and track 20+ faces simultaneously. It depends only on hardware.
7
Face movement
Face fitch, yaw and movement recognition in the timeline.
8
Gender recognition
9
Age recognition
10
Smile detection
11
Mood recognition
12
Ethnicity recognition
1
Only real faces
Algorithms trained against faces spoofing (faces on POS materials, video walls, other screens, T-Shirts and so on).
2
Mask detection
Faces in masks detection support. Do not "loose" your audience even in COVID time.
3
Staff exclusion
Separate staff from statistics data. Staff will be marked to exclude it from the main analytics data. No selling ads content to them if you don't want it.
4
Hair detection
Facial hair and hairstyle and hair colour recognition.
5
Accessories detection
Glasses and headwear recognition.
6
Group recognition
No faces count limits. The algorithm can recognize and track 20+ faces simultaneously. It depends only on hardware.
7
Face movement
Face fitch, yaw and movement recognition in the timeline.
8
Gender recognition
9
Age recognition
10
Smile detection
11
Mood recognition
12
Ethnicity recognition
Conversion in real views
Information about companies conversion rate filtered by devices. Views count: Displayforce detects people immediately (during their standing in front of device playing content) with 98%+ accuracy (the distance depends on camera resolution).
Audience portrait
Information about audience demographics filtered by devices. Gender classification with up to 90% accuracy. Age classification: 3 age groups (child, youth, adult), absolute age (+/- 5 years).
Engagement (Dwell-time)
A/B testing: dwell-time and audience data comparison between various companies. It`s necessary to understand which content has mostly influenced on the target audience. You can be sure that you're playing your advertising content in the right place, at the right time, in front of the perfect audience.
Privacy compliant
By default no sending photos, videos etc from device to comply GDPR & CCPA.
Local computations without server and transferring video streams or images. Work without extra devices and traffic out of the box.
Algorithms work locally on the devices without an internet connection. No servers needed for computer vision. No server and traffic costs.
Collecting only impersonal data: audience views, emotions and behaviour for insights.
Quality
Support
Aggregated data
Customer
Journey
Our algorithms used to "know" audience or "know" their previous movements and behaviors without personal data.
Ability to "know" the same audience in next appearing in front of the different device.
Exchanging anonymous audience hashes between devices to cluster analysis of all audience inside the POS, know their movements, behaviours, interests and use it to show relevant messages.
Customer journey algorithms create a path of all audience in POS with their attributes, interests, mood changes etc.
Impersonalized match
P2P synchronization
Customer heatmap
Case studies
Computer vision and machine learning algorithms intended for advertising and digital signage purposes.
Major features
Top-ranked
Algorithms in the top of NIST Computer Vision rank list. One of the best in speed and accuracy.
Fast response
Very fast algorithm in real cases usage. Up to 300 ms for age/gender detection and up to 1 second to recognize attributes. Less than a second to show the relevant message to a customer.
Mass devices
Uses classic USB RGB cameras. No special hardware requirement. Easy and cheap installation.
Real-time
Trigger an audience right now without delays and remote processing.
Face detection & match
With faces pitch/yaw and person movement, our algorithm knows that the face is the same.
Low quality Face recognition
Own neural network and data sets trained to best suit for real cases (not ideal "passport-like" photos with good light and so on).
Real-time triggers
Content triggering by any attribute, fact of presence, social demographic features etc to influence your audience in the moment of their choice.
Optimized
For low-end devices with simultaneously content playback. Different CPUs support x86, 64-bit, ARM, ARMv8.
Top-ranked
Algorithms in the top of NIST Computer Vision rank list. One of the best in speed and accuracy.
Fast response
Very fast algorithm in real cases usage. Up to 300 ms for age/gender detection and up to 1 second to recognize attributes. Less than a second to show the relevant message to a customer.
Mass devices
Uses classic USB RGB cameras. No special hardware requirement. Easy and cheap installation.
Real-time
Trigger an audience right now without delays and remote processing.
Face detection & match
With faces pitch/yaw and person movement, our algorithm knows that the face is the same.
Low quality Face recognition
Own neural network and data sets trained to best suit for real cases (not ideal "passport-like" photos with good light and so on).
Real-time triggers
Content triggering by any attribute, fact of presence, social demographic features etc to influence your audience in the moment of their choice.
Optimised
For low-end devices with simultaneously content playback. Different CPUs support x86, 64-bit, ARM, ARMv8.
All components in one Platform
Platform takes a unique approach, employing ground-breaking technological innovations to barriers that untill now have made digital out-of-home a grid and time-consuming medium to manage.