Ing. Jelle Meeus
Software Developer

Thesis - Continuous unobtrusive user authentication using gait for wearable devices, utilising machine learning algorithms

Text PDF Powerpoint Slides PDF

I researched a gait-based authentication method amidst growing interest in non-explicit user input. I enhanced an existing barebone implementation, featuring an Android wearable app for data recording and a Python server app for offline processing. I integrated human activity and gait recognition systems, employing traditional machine learning models and a novel, fast, and accurate feature extraction technique. The outcome is a seamless, continuous gait-based authentication system, enabling offline data capture, server-based training, and real-time evaluation on a wearable.

8051 microcontroller instruction set IEEE754 32bit Floating-Point Library

Implementation of an IEEE754 Floating-Point library for aduc832 system platform for 8-bit 8052 based systems. This library allows for accurate and fast calculations of the four basic operations (+, -, *, /) for two numbers in IEEE754 32-bit floating-point format without using MUL/DIV instructions

Software Engineer

Sioux Embedded Systems Nov 18 - Feb 19

I wrote tools accommodating a customer's migration from a Perforce Version Control System (VCS) to Git. I implemented existing and new features using Python and Gitlab API to meet developer demands. I setup a R&D Internal website (Bootstrap) to provide an overview of projects that updates nightly. I did smaller IT related tasks, such as setting up automated backups of firewall settings and Jenkins configurations. Also, I wrote some automated tests in an in-house testing framework

Machine learning algorithms student

OneSpan 4 weeks summer 2017

Developed a demo allowing recording and analysis of gait data on an Android wearable device