3:00 pm - 4:00 pm
Abstract: Humans naturally use their hands to interact and communicate with their surroundings. Reconstructing these complex and dexterous hand interactions enables sign-language recognition and translation, better assistive robots, and more immersive human-computer interaction (e.g. for AR and VR). To make hand reconstruction usable for the aforementioned applications and to a wide set of users, the methods need to work accurately and robustly in real time and with a lightweight and affordable hardware setup. In this talk, I will discuss some of our recent projects that pushed the state of the art towards lightweight real-time hand reconstruction: hand tracking in presence of strong occlusions and clutter, hand tracking from a single RGB camera, and simultaneous pose and shape reconstruction of two interacting hands.
Bio: FranziskaMueller is a 4th year PhD student in the Graphics, Vision and Video Group headed by Prof. Christian Theobalt at the Max Planck Institute for Informatics in Saarbrücken, Germany. During her PhD she has also worked at Stanford University with Prof. Leonidas Guibas and at Facebook Reality Labs Pittsburgh with Yaser Sheikh. Prior to her doctoral studies, she obtained a M.Sc. (2016) and B.Sc. (2015) in Computer Science from Saarland University in Germany. Her research focuses on reconstruction and tracking of hands in challenging scenes, in real time, and from a lightweight hardware setup. She is interested in model- and optimization-based techniques as well as machine learning for computer vision.