My research topic is Gesture Detection for Sign Language Interpretation using Wearables and deals with the question of how to reliably recognize gestures while avoiding external observers, such as cameras, for privacy reasons. The approach to gesture recognition should be mobile and able to return qualitative feedback to the user. For this purpose, different parameters of the gesture have to be distinguished, e.g. hand shape, movement, place of execution of the gesture and orientation of the hand.
As a result of my research, I must address the following questions:
- How do individual sensors impact recognition of hand shape?
- How to determine the relative position of the performed gesture?
- How to reliably determine a gesture’s start and end in continuous data?
In the context of my research, I am concerned with the following topics:
- Machine Learning
- Sensors & Sensor Fusion
- (Serious Games)
Through my affiliation with the serious games group, I always try to see my research in this context as well.