Poland |April 8, 2026
Students from Warsaw University of Technology achieved first place at the national SignAI hackathon with an innovative gesture recognition system powered by machine learning.The winning team, representing the Radar and Digital Signal Processing Student Research Group from the Faculty of Electronics and Information Technology, developed a project focused on gesture analysis using a Time of Flight sensor and a random forest algorithm.Their system uses a custom-built data acquisition and processing architecture, achieving an impressive classification accuracy of 97.49 percent. Instead of traditional Fourier transform methods, the team introduced a novel approach by calculating the discrete derivative of distance over time. This significantly reduces computational demands, enabling advanced AI algorithms to run directly on microcontrollers.The solution also incorporates a virtual radar concept and utilizes VCSEL technology to improve signal precision by minimizing interference. These innovations allow for efficient and accurate gesture detection without physical contact.Beyond its technical achievements, the project has strong real-world potential. The technology can be applied in Internet of Things environments, particularly to support individuals with disabilities by enabling intuitive, touch-free control of devices.The team members behind the project are Bartosz Falęcki, Kuba Hałaczkiewicz, Łukasz Łagosz, and Maksymilian Zabierowski, who completed the solution during an intense 24 hour coding session.
Source: Warsaw University of Technology
#WordMain #StudentNewsPortal #Europe #studentnews
About the Author
