AI-Powered Module

Professional Face Recognition
by Xeoma for personal
and business uses

Real-time face detection and recognition directly on your server.
No cloud dependency. No per-image fees. Accuracy up to 99%.

Face recognition in action
Face detected
90%
Average accuracy
Real-time
Live video stream
Local
No cloud, no bandwidth waste
3 modes
Flexible pricing and accuracy
What it’s used for

One module. Dozens of scenarios.

This is how Face Recognition module icon looks like Face Recognition in Xeoma works across industries — from finding a missing person in a city-wide camera network to recognizing a VIP customer the moment they walk into a hotel.

Face recognition in action
Access Control
Automated entry for employees and authorized visitors
Offices, hospitals, kindergartens, military facilities — the gate opens only for faces on the list. Can be used to create video intercoms.
Retail VIP recognition
Retail & Hospitality
VIP customer recognition
Notify staff the moment a key customer walks in. Also: detection of queues, unique visitors counting.
Police face recognition
Law Enforcement
Search for missing people across city cameras
Home stranger alert
Home Security
Alert only on strangers — not family members
Alternative to Motion Detector in difficult conditions.
Hospital staff tracking
Healthcare
Staff check-in tracking, mask detection, restricted access
Work time tracking
HR & Productivity
Automated work time tracking — no badge readers
Face blurring
Privacy Masking
Mask faces for privacy reasons
How it reacts

Four ways to configure what triggers the system

Face Recognition doesn’t just detect — it decides. Tell it exactly whose appearance should matter and whose shouldn’t.

React to any person
Mode 1
React to any person detected

The system alerts whenever any face appears in camera view — regardless of who it is. Best for restricted areas where no one should be present after hours: server rooms, warehouses, kitchen areas after closing time.

Use: empty zones, after-hours security, restricted access areas
React to strangers only
Mode 2
React only to strangers

Build a database of known people — employees, family members, regular visitors. The system stays silent when they appear and triggers only for unrecognized faces. Eliminates false alarms from your own staff.

Use: offices, homes, retail back offices, warehouses with regular staff
React to specific people
Mode 3
React only to specific known people

Add a list of specific individuals and trigger only when they appear — ideal for tracking work hours automatically. The camera logs exactly when each employee clocks in and out, with a timestamped snapshot or searchable video footage.

Use: attendance tracking, VIP recognition, parole monitoring
React to masked or unmasked people
Mode 4
React to masked or unmasked faces

The Mask Detector (part of the Face Recognition module) detects whether people in camera view are wearing a face mask or not — useful for compliance enforcement in medical facilities, construction sites or clean rooms.

Use: hospitals, pharmacies, food production, construction safety
Why Xeoma’s face recognition

Built differently from the rest

Most face recognition solutions send your footage to external cloud servers. Xeoma processes everything locally — meaning your data never leaves your building.

Local processing — no cloud
Safe to use

Everything stays on your machine

Your footage, your faces, your data. Recognition happens entirely on the local server — no images are sent to any external service. This also means the system works fully offline and doesn’t consume upload bandwidth.

Works on isolated networks. Zero cloud dependency. GDPR-friendly by design.
Real-time face recognition
Fast work

Real-time — not frame by frame

Xeoma processes live video streams, not individual uploaded photos. The higher the frame rate of your camera, the better and faster the recognition. Multiple faces in a single frame are handled simultaneously.

People can be standing still or moving: no movement speed limitations.
Affordable pricing
Budget-friendly

One-time payment. No per-image billing.

Many recognition tools on the market charge per image or per month. Xeoma’s face recognition is a one-time license purchase. Process as many faces as you want, forever, with no ongoing costs.

Three options to fit any budget: two Artificial intelligence options included in Xeoma Pro, and the advanced Vision method available as a separate Additional module.

“The first test I did with the facial recognition Vision algorithm was on a 360-degree camera in the hallway leading to our offices. I was amazed at how efficient Xeoma is, because this test was done on a server with only an older Intel Core i7 CPU, and still worked great! It’s not ideal, obviously, but it still worked. It was even more impressive that it worked simultaneously with the Fisheye Dewarping, which allowed us to recognize humans and faces from end-to-end in the hallway. Without the fisheye dewarping, it won’t recognize faces upside down, but with the dewarping, it recognizes them perfectly! The fact that this was all done on a CPU-only system was astounding.”

BR
Bryan Russell
Founder & CEO — Duluth Security Needs
How it works

Three recognition modes — pick the one that fits

Xeoma offers three independent algorithms for face recognition. Select one that fits to your budget and accuracy requirements.

Artificial Intelligence method
Xeoma Pro / Additional license
Artificial Intelligence
~90% accuracy
  • A Xeoma Pro license or an Additional license
  • Fast processing
  • Multiple faces simultaneously
  • Licensed per the number of faces, regardless of camera number
  • Indoor and outdoor use
  • Ignores glasses and beards
  • Search by name or photo through recordings available
  • Unique visitors counter available
  • Database sync across servers
Setup guide for the AI method >>
CNN Asian face recognition
Xeoma Pro / Additional license
AI+CNN (Asian edition)
~90% accuracy
  • Version of the AI method optimized for Asian faces
  • A Xeoma Pro license or an Additional license
  • Fast processing
  • Multiple faces simultaneously
  • Licensed per the number of faces, regardless of camera number
  • Indoor and outdoor use
  • Ignores glasses and beards
  • Search by name or photo through recordings available
  • Unique visitors counter available
  • Database sync across servers
See setup guide for the AI method >>
CSV reports: Face Recognition can save recognition events directly to CSV spreadsheet files — import into Excel, LibreOffice or MySQL to build any attendance, access or analytics reports you need.

No special hardware needed

The free edition is a genuine product — not a crippled demo. Up to 1,000 cameras, no time limit, no watermarks.

1
Any computer, standard specifications

Regular off-the-shelf cameras and computers are all you need. Xeoma handles the recognition processing on its side.

2
Any camera with a clear image

No embedded recognition detectors or special lenses needed. What matters is image quality and camera angle. Point the camera as perpendicularly as possible toward faces, and ensure faces are visible for at least a few seconds.

3
Affordable price

Xeoma has flexible pricing with its prices coming as monthly subscriptions or one-time-payment licenses. Affordable competitive costs.

Open requirements calculator
Watch it in action

More about Face Recognition, scenarios, and full setup

See what Face Recognition can do and get a complete step-by-step configuration guide and examples of usage.

Quick start

Test it in 5 minutes

1
Launch Xeoma in Trial or Pro mode

Download and run Xeoma — it starts in Trial edition automatically, which lets you test nearly all features including Face Recognition.

2
Add a camera

Let Xeoma auto-detect cameras on your network, or add one manually by IP address. Any types supported. Demo camera available.

3
Add the Face Recognition module to the chain

Drag it from the module list into your camera chain — between the camera source and your desired reaction (archive, notification, etc.).

4
Choose the desired method

Select a method to try. You can try all methods to find the best fit for your project. Full comparison here.

5
Train the module and test

Add a few faces to the database and see the system recognize them in the live stream. Detailed setup guide here.

6
Connect reactions

Recording, notifications, custom reactions, spreadsheet reports – select the result you need for your face recognition. More about notifications in Xeoma

Xeoma face recognition module settings
The Trial mode has several limitations. To ensure a more thorough and convenient evaluation of Face Recognition, we recommend requesting a free, non-committal demo license. This allows you to test the full range of features without restrictions. Simply click the button below to get started:
Get a Free Demo License

Ready to put faces to names — automatically?

Try Xeoma’s Face Recognition free. No time limit on the demo —
evaluate it in your real environment at your own pace.