Home Use Cases Campus & Workplace
USE CASES · CAMPUS & WORKPLACE

Campus & Workplace

“You called the roll. Does that record match who was actually in the room?”

Calling names and tapping cards verifies a voice and an object, not a person. At the classroom, office and canteen door, UniverseAI verifies the person by face and returns that result to your academic, HR and canteen systems. No new cards, no new app to hand out.

Send Your Operating Conditions Request a Live Demo

What actually happens in classrooms and workplaces

We will not open with market size. Here are the three scenes your staff live through every day.

SCENE 1 / TAKING THE ROLL
Checking who is here eats into class time
The larger the class, the more of the opening minutes go into calling names. And even then, someone answering on another student's behalf is not caught in the room.
Cause: calling a name verifies a voice, not a person.
SCENE 2 / CARD-BASED ATTENDANCE
A card can be handed to someone else
A time record tapped with a badge is, strictly speaking, a record that the card was there at that moment. Once a lent badge works even once, every time record after it loses its footing.
Cause: what is being verified is not the person but the object the person is holding.
SCENE 3 / CANTEEN HEADCOUNT
The meal count does not match who actually ate
Company and dormitory canteens check eligibility against a roster and a card. The resulting meal count drifts from actual usage month after month, so settlement and ordering rest on estimates.
Cause: eligibility is checked per roster entry and per card, not per person — so the usage log never records who actually came.
The three scenes look like separate problems, but there is only one defect — what gets verified is not the person, but the voice they produce or the object they carry. Change the thing being verified to the face, and attendance, time records and meal counts all stand on the same footing.

GATE + VCA — verify by face, retrace when you need to

Your systems record attendance, time and meals from the verification result GATE produces. Retracing movement on campus or on site after an incident is VCA's job.

A configuration diagram in which the terminals and the academic, HR and canteen systems are already yours, only GATE — performing liveness and 1:N matching — is newly installed inside your servers, and VCA sits below as a secondary lane for after-the-fact review The terminals are yours; the verification runs inside your servers, in GATE. Grey = exactly what you use today / blue = what is newly added Already yours Newly added Already yours Your terminals Classroom, door, canteen They capture the face GATE Liveness spoof detection 1:N matching Installed on your servers Your school and HR systems Attendance, time, meals Door control too Image sent Result back VCA — after-the-fact review (secondary) When something happens, combine face, body, clothing colour and carried-item conditions to retrace movement on site Attendance records, academic processing and door control are handled by your systems. Our scope ends at returning the verification result.

Only the middle stage is new. Your terminals and your academic, HR and canteen systems stay as they are — only the identity check between them changes.

Component What it does Why it is needed
GATE Inside your own servers it runs a spoof check (liveness) on the image the terminal sent, matches it 1:N against your enrolled population, and returns who the person is. A card or an app can be lent; a face cannot. Attendance and time records are kept per person.
VCA Layered onto the cameras already installed across the campus or site, it combines face, body, clothing colour and carried-item conditions to support after-the-fact review. Attendance only records the moment of the check. Retracing what happened afterwards is a separate job.
Used together At the door everything is recorded per person; inside, what happened can be retraced with a few conditions. It removes the state where the records are accurate but nothing can be checked.
“Attendance covers the moment of the check. Everything after that is VCA.”
Schools
The classroom door

Attendance is recorded automatically by face at the classroom door. Outcomes are calculated per site — send us classroom count, headcount and how attendance is currently run, and we will calculate and reply.

Workplaces
The office and site entrance

Access and time records rest on the same check. When several sites are managed together, the enrolled population grows until 1:N scale is itself the requirement.

Canteens
Company and dormitory canteens

Eligibility is checked per person instead of per roster line. The meal count lines up with actual usage, so settlement and ordering come down off estimates.

Adoption takes four steps

From confirming operating conditions to integration — here is exactly what is exchanged at each step.

1
Confirm operating conditions
Classroom count, headcount and how attendance is run today. For a workplace: number of doors, enrolled population, and how time records and meals are handled. Alongside these we confirm three things — the environment to host the authentication server, the terminals that capture the face, and who will do the integration.
2
Live demo
We run the full path from enrolment to authentication live. You see for yourself how liveness screens attempts to check in with a printed photo or a video replayed on a screen.
3
Pilot verification
Verification under the real classroom and doorway conditions. Enrolment image quality and lighting drive the result, so the verification runs under the same conditions as live operation. Scope and duration are agreed case by case.
4
Integration and rollout
We hand over the integration specification, and your development team or SI partner connects it to the academic, HR and canteen systems. Expansion means adding classrooms and doors — the server is not rebuilt.
Where our scope ends
Our part is to receive the image the terminal sends, run the spoof check, match it 1:N, and return the result. Recording attendance, opening the door and settling meal counts from that result belong to your systems. Drawing this boundary clearly is what keeps integration scope and accountability settled from day one.

What we base these claims on

Only what has been measured and what has actually been contracted — followed by the scope, stated in the same place.

99.97%
Face recognition rate under KISA certification testing
Under 1 sec
1:N matching response target
400M+
Face database scale
National university
Face recognition attendance service contract
Storage method and deployment

An on-premises installation on your own servers, storing feature data for matching rather than original face images (template-only). It can be adopted even where biometric data may not leave the institution.

Certifications and origin

A Korea-developed, non-Chinese vendor holding KISA, iBeta, GS Grade 1 and ISO certifications. If your school or institution requires a specific document format, we prepare the documents in that format.

Stated up front
  • We return the identity result; your systems record attendance, control the door and settle meal counts. This is not a product that replaces your academic system.
  • Liveness screens the common spoof attempts — printed photos, video replayed on a screen, masks. Field performance tracks camera quality and lighting, so we tune the decision thresholds to your installation environment.
  • The 99.97% figure under KISA certification was measured under accredited test conditions. Field numbers move with camera specification, lighting and enrolment image quality, so we recommend verifying in your real environment first.
  • If this describes you, we are the right fit — you have an environment to host the authentication server, an enrolled population large enough to require 1:N matching, and a rule that biometric data cannot leave the premises. Send us the scale and the environment and we will review against those conditions and reply.
  • VCA's after-the-fact review narrows the candidates; your officer makes the final determination. It is not a feature that automatically identifies a person from a single photograph.

Frequently asked questions

The nine questions school and workplace staff actually ask first.

Can attendance work by face alone, without a card or an app?
Yes. Identity is established by matching 1:N against enrolled faces, so nothing has to be carried. A terminal or app that captures the face is still required, and in most cases that is the equipment you already use. Tell us what terminals you have and we will confirm whether they integrate, and reply.
Where is the attendance record kept? Do we have to change our academic system?
No. We return the identity result, and the system that records attendance from it is the academic or HR system you already run. We hand over the integration specification and your development team or SI partner connects it.
Could someone check in for another person using a photo or a video on a screen?
Liveness screens the common spoof attempts: printed photos, video replayed on a screen, and masks. It runs on ordinary colour cameras, and the user does not have to blink or turn their head. Field performance tracks camera quality and lighting, so send us the installation environment and camera specification and we will tune the decision thresholds to those conditions and set out a verification method.
Are student and staff face images stored? We are concerned about personal data.
It is template-only: feature data for matching is stored, original images are not. Deployment is on-premises on your own servers, so the data stays inside the institution. We provide access-rights management and audit logging — tell us which regulations apply to you and we will list the required items and respond.
We have many classrooms and a large population — can it handle that?
1:N finds who a person is within a large enrolled population, and the system is built around a 400M+ face database with a 1:N response designed for under one second. The actual configuration still depends on enrolled population, daily authentication volume and concurrency, so send us those three numbers and we will review and reply. We do not confirm on the spot.
Can it also be used to check canteen or dormitory dining eligibility?
Same configuration. We match the image sent by the canteen door terminal to confirm the person is an eligible user, and return that result to your meal management system. The headcount that used to be tallied from rosters and cards is now kept per actual diner. Send us the number of canteens and daily covers and we will review the configuration and reply.
We do not run our own servers. Can we still adopt this?
This configuration suits organisations that have a data centre or cloud environment to host the authentication server. Tell us what that environment looks like — an in-house server room, a campus IT centre, or cloud — and we will review what configuration is possible under those conditions and reply.
If an incident occurs, can we also trace movement across the campus?
That is exactly where VCA sits. It layers onto the ordinary cameras already installed across the campus or site, narrowing candidates by combining face, body, clothing colour and carried-item conditions, then showing routes and companions. Cameras are not replaced. Note that it narrows candidates — your officer makes the final determination.
What does adoption cost?
We respond with a quotation. To prepare it we need the number of classrooms or doors, the enrolled population, daily authentication volume, and the type of terminals you use today. Send us those four and we will respond with a configuration.

Related products

Detail pages for the two products that make up this package.

GATE — 1:N biometric authentication on your servers →

Liveness spoof detection, template-only storage, integration specifications and FAQ
VCA — The video search brain →

Condition-combination search on face, body, colour and carried items, plus companions and routes

So the record is kept per person

Send us the number of classrooms or doors, the enrolled population, and how attendance and time records are run today — we will work out a possible configuration and respond with a demo schedule.

Send Your Operating Conditions