Arx Bank – Facial Recognition Banking App

Arx Bank is a cutting-edge banking application developed for a hackathon with the theme of facial recognition. The project won first place by demonstrating innovative use of AI and machine learning in financial security.

The application uses a Python-based facial recognition system powered by the K-Nearest Neighbors (KNN) algorithm. During registration, users are guided to move their head around while the system captures multiple photos from different angles. These images are then used to train the machine learning model on-the-fly, creating a personalized face profile for each user.

Every login attempt requires facial recognition authentication. The system handles multiple edge cases intelligently: it detects when no face is present in the camera frame, prevents authentication when multiple people are detected, and ensures only the registered user can access their account. For added security, sensitive features within the app also require face verification, and the application automatically logs out if an unauthorized person is detected in front of the camera.

The application was built using Tauri and React, leveraging Rust's performance and security features to create a standalone executable that can run on Raspberry Pi devices.