Context Recognition Seminar


Tutor: Marcus Handte
Nowadays, most devices are equipped with differnt types of networking technology and a broad spectrum of sensors. Examples include gyroscopes, accelerometers, cameras, and microphones. To provide better task support, future applications will have to use the sensors of multiple devices to determine the state of their environment in an automated fashion. This requires novel software systems and signal processing algorithms to derive high-level context information from low-level sensor readings.

The seminar topics will cover a selection of systems and algorithms to recognize different types of context. Furthermore, they cover supportive architectures and protocols to recognize context in a distributed manner. Thereby, the seminar focuses on light-weight approaches that can be implemented on resource-poor devices. Participants will be able to select their topic of choice from the set of available topics.

This seminar is suitable for students at the master level. Participants will have to do a literature research and they will have to create a written high quality report. Furthermore, they will have to give an oral presentation of their topics. Note that it is not possible to combine this seminar with the context recognition with mobile devices project. Please also note that the maximum number of participants is limited. Depending on the number and type of participants, this seminar might be given in English and German. If you want to participate in the seminar, please send an email to marcus.handte@uni-due.de.

The kickoff meeting for this seminar will take place on Thursday, October 25th (moved from 18th due to conflict with student plenary meeting) between 13.00h and 15.00h in BC504. Participation in this meeting is mandatory.