Robert Pauls

Delve into a vast range of Formula One statistics from driver's points to the sport's circuit history, and create your own insights, all using data available since the first championship.

YOLO-based object detection pipeline trained on a custom dataset to track Formula 1 cars in static and real-time footage, achieving over 90% accuracy across varying conditions.

Phishing classifier using CBOW embeddings trained on 18,000+ Enron emails, achieving 96% accuracy. Integrated persuasion-pattern detection alongside semantic classification.

Multiclass computer vision system comparing three CV pipelines to classify face mask status across 2,800+ images, with a custom CNN reaching 91% accuracy against real-world 'in the wild' testing.

A PHP MVC application with custom controllers and models handling user auth, volunteer sign-up, and staff shift scheduling, containerised with Docker Compose on a Linux-based LAMP stack.

A robust detection system for motorsport vehicles in mainstream racing categories including Formula, Prototype and GT3.