Face photo vs face recognition attendance for contract staff
Face photo attendance means a supervisor taps a staff member's photo on a phone to mark them present. The supervisor is the check, and no face data is processed. Face recognition attendance means a camera scans the person's face and software matches it against a stored face template. For contract staff spread across many client sites, face photo marking needs no enrolment scans and no extra hardware, and it works offline. Face recognition removes the supervisor from the check but stores biometric templates.
By HaazriPro Team · Last updated
Competitor facts checked on 8 October 2026. Each one links to the vendor's own page.
How does face recognition attendance work?
A face recognition system first enrols each staff member: it captures their face and turns it into a mathematical template. At marking time, a camera (a phone, a tablet at the gate, or a dedicated device) captures the face again and the software compares it with the stored templates. If it finds a match above its confidence threshold, it marks that person present.
Vendors selling to Indian employers advertise this today. For example, PagarBook describes "face recognition at the gate" with a stated "99.2% accuracy" on its home page, and says on its pricing page that its Lens face recognition "works offline". SalaryBox lists "AI Selfie Attendance" on its home page and includes face recognition in its Business and Enterprise plans on its pricing page. Accuracy figures are the vendors' own claims; we have not tested them.
How does face photo attendance work?
In HaazriPro, each staff member's photo is taken once, when they are added. At the site, the supervisor's phone shows a grid of face photos for the staff assigned to that site and shift. The supervisor looks at who is actually in front of them and taps their photo to mark present, late, half-day, absent, leave, overtime or transferred. A coloured ring on each photo shows the status.
There is no camera scan at marking time and no matching software. The photo is there so the supervisor picks the right person, which is the problem with name-based registers: Sunil, Sushil and Suresh are easy to mix up on a list, and hard to mix up by face. Marking works fully offline and syncs when the phone gets a signal.
How do the two methods compare?
| Face photo (HaazriPro) | Face recognition (category) | |
|---|---|---|
| Who confirms identity | The supervisor, by looking at the person and the photo | Software, by matching a face scan to a template |
| What is stored | One ordinary photo per staff member | A face template derived from each person's face, plus usually a photo |
| Enrolment | Take one photo when adding the staff member | Enrol each face; re-enrol if matching fails |
| Hardware | The supervisor's Android phone | A phone, a tablet at the gate or a dedicated device, depending on the vendor |
| Who must be present | A supervisor at each site | Each staff member in front of a camera |
| Offline | Fully offline, syncs later | Depends on the vendor (some state that it works offline) |
| Proxy marking | Hard: the supervisor marks the face in front of them | Hard: the camera must see the person |
| Weak point | Relies on an honest supervisor | Relies on the camera, lighting and the vendor's matching model |
Which one suits contract staff at client sites?
Contract staff work where the client is: a housing society gate, a mall basement, a factory floor, a construction site. There is usually one supervisor responsible for several posts, the staff rotate between sites, and many sites have weak or no mobile network. That shapes the choice:
- Sites change often. With face photo marking, adding a new site means assigning staff to it in the app. A fixed device has to be bought, installed and powered at every new site.
- Staff turnover is high. A new joiner needs one photo, taken in seconds. Face recognition needs an enrolment that works reliably for that person.
- The supervisor is already there. On most contract sites a supervisor checks the posts anyway. Face photo marking makes that check the attendance record.
Face recognition makes more sense where many staff pass one fixed entrance, such as a factory gate or a warehouse, and the employer does not want any person to vouch for another.
What about your staff's data?
A face template is biometric information about each staff member. Before you collect it, you should know what you are collecting, why, how long you keep it, and who can see it, and you should tell your staff. Laws on personal data apply to employers as well as to software vendors, so take advice from your lawyer before rolling out any biometric system.
Face photo marking keeps the record simpler: you hold an ordinary photo of each staff member, the same kind of photo you already put on an identity card. HaazriPro stores the photo compressed in private cloud storage, and every record is limited to your own agency's account.
Does face photo marking stop proxy attendance?
It stops the most common kind: one staff member marked present on another's row by mistake, or a friend asking for a mark on the register. The supervisor has to tap a face, so they have to see the person. It does not stop a supervisor who deliberately marks someone who is absent. HaazriPro records when each mark was made, so the owner can see when a day was back-filled later. Read more in our guide on why name-based attendance causes pay disputes.
Can you combine the two?
Yes. Some agencies use a fixed face recognition device at one large site with a single entrance, and phone-based marking everywhere else. What matters at month end is that all attendance ends up in one place, so that salary and client bills are worked out from the same record.
How do the costs compare?
Face recognition is usually priced as an add-on per staff member or as a device per site; check each vendor's current pricing page. HaazriPro is one flat price of ₹499 a month for up to 500 staff, with unlimited sites and supervisors, and runs on the Android phones your supervisors already carry.
Sources
- PagarBook home page (checked 8 October 2026)
- PagarBook pricing (checked 8 October 2026)
- SalaryBox home page (checked 8 October 2026)
- SalaryBox pricing (checked 8 October 2026)
This comparison reflects each product's public pages as we checked them on 8 October 2026. Products and prices change, so check each vendor's current pages and take advice from your lawyer on staff data.
Try face photo attendance on your own sites
Supervisors mark attendance by tapping each staff member's face photo, even offline. The salary sheet and the client bill are ready at month end.
Start 30-day free trial