System Failed
human path prediction - interactive ML application with GUI
Overview
System Failed is an AI-powered interactive installation that combines advanced motion tracking, machine learning, and real-time audience interaction. I developed the entire software pipeline, from processing raw laser tracking data to generating live path predictions. Key features include the ability to record and retrain models using new data and a PyQT5-based graphical user interface for system control. The production was succesfully shown to a live audience 11 times.
GitHub Repository | Project Page | Technical Page
