Artificial Intelligence Developed by University Students Recognizes People by Their Gait

A new artificial intelligence-based system developed by university students can identify individuals in real time with a high accuracy rate using three-dimensional data obtained from walking movements.

Working Principle of the System

This developed artificial intelligence technology works by training pre-collected data on the model, much like commonly used facial recognition systems.

Before the system can recognize people in real time, it undergoes a specific data collection process from the user.

Point Cloud Technology

While processing the point cloud obtained from human movements, the system sends waves within a specified frequency range and collects the returning waves.

These collected waves are converted into a point cloud for each second, and points are taken from various parts of the body, including the head and shoulders.

3-Dimensional Coordinate Analysis

These captured points are recorded in the system as three-dimensional X, Y, and Z coordinates of individuals in space.

Thanks to these precise measurements and the point cloud method, individuals are uniquely identified in the digital environment.

Success and Accuracy Rates

According to information provided by student Betül Nuriye Koparmal, a 93% success rate was achieved in evaluations conducted on a ready-made dataset of approximately 15 people.

The researchers stated that they achieved an accuracy rate between 85% and 90% on the data they collected themselves.