Research

Project by Yujun Piao

As a born-to-be computer enthusiast, I have always been interested in using engineering and machine learning to study real-world problems more directly.

I built PressurePoint, an IoT wheelchair system that detects tilt, posture, and falls; Project Signify, a 99%-accurate ASL/KSL recognition model built from a custom dataset; and Dalseong AI, a 97%-accurate model classifying 10 species’ calls for real-time biodiversity monitoring at Dalseong Wetland. Alongside these projects, I have also conducted comparative energy-output tests for SpiralPanels, treating research not as a paper alone, but as a cycle of field data, prototype, validation, and public use.