Real-time facial recognition
AI-powered recognition identifies faces as they appear, so attendance is captured live using ArcFace embeddings and FAISS similarity search.
attend is a Windows desktop application that uses AI-powered facial recognition to manage attendance in real time.
Application screenshot placeholder
Add the real screenshot as assets/img/app-screenshot.png. It will appear here automatically.
Privacy-focused, offline-first attendance management with advanced facial recognition technology.
AI-powered recognition identifies faces as they appear, so attendance is captured live using ArcFace embeddings and FAISS similarity search.
Two independently configurable daily sessions with automatic duplicate prevention and manual override capabilities for administrators.
Operates entirely on your local machine without cloud connectivity. All student records and biometric data remain under your control.
Enroll students individually or in bulk from a directory. Supports JPG, PNG, and BMP formats with automatic FAISS index updates.
No GPU required. Optimized for CPU-based processing with ONNX Runtime, making it accessible on standard hardware.
Optional biometric data encryption, seven-day unknown face retention, and administrative action logging for compliance.
Designed to run on standard Windows hardware without dedicated GPU requirements.
Operating system
Windows 10 / 11
Processor
CPU with AVX support preferred
Memory
4 GB RAM minimum
Storage
1 GB free space
Camera
Webcam required for live recognition
Get attend running on your Windows PC in a few steps.
Select Download for Windows to save the .exe file to your computer.
Open the downloaded file and follow the on-screen prompts. The app installs to your user directory under Programs.
Open attend when setup finishes. A Start Menu shortcut is automatically created.
On first launch, the InsightFace recognition model downloads automatically. An internet connection is required only for this initial setup.
One installer, ready to run on your Windows PC.
Checking installer availability…