productivity

Pop-up Face Check-in

A local-first face-matching check-in tool for temporary events, with manual check-in and entry/exit tracking when needed.

View on GitHub

Local processing

About your filesFiles and content selected in this tool are processed in your browser and are not uploaded to a Browser Kitty backend.

Operation demo See the basic workflow in a short video.

About this tool

Pop-up Face Check-in is designed for temporary reception desks at events and gatherings. Participants can be registered in advance, then the browser detects faces, extracts matching features, and compares them locally during check-in. If matching fails, staff can search by name and check the participant in manually.

It supports either a one-time check-in or a chronological entry, exit, and re-entry mode. Face-matching data stays on the device and is excluded from CSV/JSON attendance history. The workflow is intentionally temporary: face-related data can be deleted together after the event, while password-protected .popupface files are available when registration data must be moved to another device.

Good for

Speed up temporary reception for internal events, meetups, workshops, or gatherings
Register a participant list in advance and reduce manual interaction at the desk
Keep manual name-based check-in available when face matching fails
Record entry, exit, and re-entry history instead of only a single arrival time

What it can do

Key capabilities available in this tool.

On-device face detection, embedding, and matching with YuNet / SFace
Name search and manual check-in when face matching is unavailable
Single check-in or entry, exit, and re-entry history modes
Multiple participants with up to three matching samples per person
Password-protected .popupface files for moving registration data between devices

How to use

Open the tool, choose Check-in or Entry/Exit mode, and select how long face data should be retained.
Register participants and optionally add up to three matching samples per person.
Start the camera at reception; repeated matches for the same person trigger automatic check-in or entry/exit recording.
If matching fails, search the registered participant by name and check them in manually, with cancellation and Undo where available.
Export attendance history as CSV or JSON, then delete all face-related data after the event when it is no longer needed.

Supported

Choose between one-time Check-in mode and Entry/Exit mode with re-entry history.
Store up to three matching samples per participant and combine automatic matching with name search and manual check-in.
Registration data can be saved to a password-protected .popupface file separately from attendance history and loaded on another device.
Camera use requires a secure context such as HTTPS or localhost and browser camera permission.

Limitations and notes

Face matching can produce false matches or misses, so the workflow should retain human review and manual check-in.
Accuracy varies with lighting, camera quality, pose, expression, registration images, and threshold settings.
It is not intended for legal identity verification, security gates, payment authentication, or other high-assurance identity uses.
Face images and matching data can identify people. Confirm the purpose, consent requirements, and organizational rules for your setting, and delete data when it is no longer needed.

Frequently asked questions

Are registration photos or camera video uploaded to a server?

No. Detection, feature extraction, and matching run in the browser. Source registration images are not retained after processing, and camera frames are not stored.

What happens if face matching does not work for someone?

Staff can search the registered participant by name and check them in manually, so reception can continue even when matching is unreliable.

Can I use it for identity verification or security authentication?

It is not recommended for that. The tool is intended for temporary event reception and attendance, not legal identity verification or high-assurance access control.

Can I delete face data after the event?

Yes. The delete-all action removes registered participants, representative images, matching data, and check-in state from browser storage. Delete exported .popupface files separately when they are no longer needed.

Offline version and source code

The standalone distribution bundles ONNX Runtime Web, WASM, YuNet, and SFace models. It can run from a local HTML file for offline use, subject to browser secure-context rules for camera access.

View source on GitHub