Computer vision-based multi-target tracking fails in especially unmanned aerial vehicle (UAV) videos due to several reasons such as occlusion and low resolution.This project aims to be able to continue tracking, when the computer vision-based tracking fails in UAV videos, by using a brain-computer interface (BCI). Hence, our goal is to develop a hybrid tracker utilizing computer vision and BCI together. The BCI component will be an end-to-end video target marking/tracking electroencephalography (EEG) system, where the stimulation is based on Steady-State Visual Evoked Potentials (SSVEP). The planned work includes 1) the stimuli design, 2) EEG experiments, 3) algorithm design and 4) performance analyses. The students will start working in this order and can continue to the very end if they wish. However, our main expectation in this PURE phase of the project is first to complete the stimuli design and the EEG experiments.
Faculty Department of Project Supervisor:
Faculty of Engineering and Natural Sciences
Number of Students:
Related Areas of Project:
Computer Science and Engineering