Two students from PPU presented and discussed the first master's dissertations on Intelligent Systems Programs. The programs comprise data science and robotics. During the event, Elias Muhammad Maharmeh presented his master's thesis on robotics, titled "Reinforcement Learning-Based Human-Machine Co-Adaptation via Policy Gradient." Another student, Islam Ayman Emar, presented his master's thesis on data science, titled "Using Deep Pose Estimation of Football Player Body for Virtual Reality."
The thesis written by Elias presents a model that proposes a solution to the problem of human-machine interaction. The model aims to facilitate joint activities between humans and machines by enabling the machine to learn, adapt, and modify its behavior. To achieve this, a new co-adaptive policy gradient algorithm, which is a type of reinforcement learning algorithm, has been proposed. This algorithm allows the machine to adjust and synchronize its behavior with that of humans during common tasks, ultimately resulting in optimal performance.
Islam's thesis focuses on developing a new approach for creating 3D models of football players during matches. The technique relies on constructing a series of deep neural networks that are trained beforehand on the players' 3D information. This provides a cost-effective alternative to capturing football games with a limited number of cameras in the stadium and enables the matches to be viewed through a 3D viewer or virtual reality devices.
Following a discussion of the two theses, the committee recommended that Elias be granted a master's degree in smart systems with a focus on robotics. Similarly, the discussion committee suggested that Islam Omar be awarded a master's degree in smart systems with a concentration on data science.
Public Relations and Media Department
