Knock Unlock

Knock Unlock is a project that uses machine learning to authenticate users by matching a pattern of knocks recorded in real time to a secret password pattern.
The program is designed to be a more convenient way to interact with home automation devices.

How it works

The process involves recording two audio files, one serving as the password, and the other as the user's input. These files are then compared using neural networks and feature extraction techniques. If the files match, the user is granted access, signifying that the password was correctly input.

Demo

_{Wrong password being input into the program (password pictured on the left) Right password being input into the program (password pictured on the right)}