TheLearningLock

An Android app that uses an anomaly detection (based on a Normal Distribution) to learn how the user enters their pattern and detects it there is an imposter

Setup

• Install JDK
• Install Android Studio + Android SDK
• Create / Get the app/src/main/res/values/secrets.xml file which contains API keys. The format should be

<?xml version="1.0" encoding="utf-8"?>
<resources>
    <string name="AWARENESS_API_KEY_DEBUG">{Google API Key for debug builds}</string>
    <string name="AWARENESS_API_KEY_RELEASE">{Google API Key for release builds}</string>
    <string name="FABRIC_API_KEY_DEBUG">{Crashlytics API Key for debug builds}</string>
    <string name="FABRIC_API_KEY_RELEASE">{Crashlytics API Key for release builds}</string>
</resources>

• Build and Run on Android Studio

Details

Minimum API - Android 4.4 (Level 19) / Kitkat

Source Code package structure

• lockscreen: Contains code related to the Lock screen that is shown when the phone is locked and any related code, such as the notifications system
• main: The Activity that is actually launched when the app starts, and the Welcome and Settings fragments
• pattern: Shared files that are used for anything related to the Pattern unlock
• pin: Shared files that are used for anything related to the PIN unlock
• setup: Code related to the Setup flow (which is used by the stuff in main)
• shared: Code that is shared by the other packages

Credits

Created by Alston Lin with contributions from Ejaaz Merali, and Tyler Wang

Designed by David Liu