Dr Xin Yuan

Academic Qualifications

Current Roles

• Research Scientist
  To develop both AI-based algorithms to process imaging and sensor data for 3D object detection and tracking, seizure detection and prediction, and other applications. • Developed an AI algorithm for detecting and tracking 3D deformable objects (meat) in motion from 3D videos. • Developed a graph neural network (GNN)-based detection algorithm for Spike and wave discharges (SWDs) detector to detect seizures based on EEG data. • Developed a Deep Reinforcement Learning (DRL)-based control algorithm (Deep Deterministic Policy Gradient (DDPG)) to autonomously and visually track the moving objects (such as UAVs and vehicles) based on the onboard video cameras. • Proposed a DRL-based beam tracking framework to learn the positions/velocities of a moving object and an appropriate beam steering angle. • Developed a computer application program that can process SRTM digital elevation and water body data (provided by USGS) to produce radio coverage maps of FM broadcasting services in Australia.

Professional Areas

• Deep Reinforcement learning
• Graph Learning
• Federated Learning
• Applied machine learning

Related links

• Google Scholar