The Perceptual User Interfaces Group at the Max Planck Institute for Informatics invites applications for fully funded PhD positions in computer vision and human-computer interaction with a focus on egocentric computer vision and mobile (gaze-based) human-computer interaction.
Wearable devices, such as egocentric cameras and head-mounted displays, are increasingly important as a research platform in both computer vision and human-computer interaction. However, despite significant advances, it is still challenging for computers to understand visual scenes and human activities from an egocentric perspective. It is also challenging to provide natural and efficient means for users to interact with body-worn devices. Because egocentric computer vision is closely linked to human vision, mobile eye tracking is a promising extension to egocentric vision systems. Visual behaviour naturally reflects users’ visual attention and perception and provides valuable information about users and how users interact with each other as well as the environment.
The proposed PhD projects will develop computational methods for gaze estimation and scene understanding in egocentric settings as well as techniques and methods for gaze-based human-computer interaction. The studentships are embedded in an international research project funded by the Japan Science and Technology Agency (JST) and conducted in close collaboration with the University of Tokyo (Japan), Keio University (Japan), and Carnegie Mellon University (USA). Visiting positions at partner institutions during the PhD project are possible and highly encouraged. The projects focus on developing systems that support users in their daily lives by utilizing large-scale multi-user visual sensing. Potential applications include real-world navigation and medical augmented reality systems.
The research will be experimental, using portable eye tracking equipment and egocentric vision systems, and will involve controlled user studies as well as data collection in mobile daily-life settings. In addition to experimental skills, the work will require to develop a thorough understanding of human vision, the design and implementation of prototype systems and interfaces, as well as statistical data analysis techniques.
The research directions of particular interest include but are not limited to:
- Robust and calibration-free methods for gaze estimation in mobile daily-life settings
- Scene and activity understanding from egocentric video
- Gaze-assisted computer vision for human-centric scene understanding
- Information visualization techniques for large-scale gaze recordings
- Interaction models and techniques for multi-user egocentric vision systems
The Perceptual User Interfaces Group at the Max Planck Institute for Informatics offers a highly collegiate and stimulating environment for doctoral research training. The successful candidate will join an ambitious research group that is at the forefront of these emerging research areas. The candidate will be expected to contribute to the strong profile of the group by participating in the preparation and publication of research results at the level of international excellence in computer vision and human-computer interaction.
We invite applications from enthusiastic individuals, who are able to work independently and have an excellent first degree in Computer Science or a related field relevant to the proposed research, and very good knowledge of computer vision or human-computer interaction. The studentship is not restricted by nationality. Interested applicants are advised to consult previous work done by the group. Applicants are also strongly encouraged to initiate contact with Dr. Andreas Bulling prior to their application.
Applicants should submit their CV, a copy of school and university degree and course transcripts with grades (Abitur, Vor- and Hauptdiplom for German applicants), names and contact information of two references and a short description of research interests. Applications should be emailed to
Incomplete applications will not be considered. Applications are accepted until all positions are filled.