Bachelor Proiect Question (Gait analysis using computer vision)

Follow the full discussion on Reddit.
Hello, i am doing my bachelor project this semester and i have been assigned this topic (Automated Gait Analysis System Using Computer vision). here's the description: - - - Research has shown that people have walking patterns (gait) that can be used to identify them with good accuracy. This topic investigates the possibility of developing a gait-based authentication system. Here, a camera will capture a person walking and pass the frames to a system that will extract features that can be used to authenticate this person. I decided to use the xbox kinect with the jetson nano since i've seen some projects, and thought the depth sensing might become handy later in my project. https://data.mendeley.com/datasets/svx74xcrjr/3 I found this dataset but I have no idea whether the data from my kinect can be similar to this one. Since i have very limited experience on machine learning and almost none in computer vision: How do you think i should start on this topic? If you have any experience using the kinect with the jetson or any ai applications, do you think it is a good choice or should i stick to regular cameras? Is the dataset attached usable or should make my own dataset whether on people with gait abnormalities or i could mimic the abnormalities which is an option? Any tips would be greatly appreciated. thank you for your help.