The focus of this seminar is on motion synthesis for virtual characters acting in virtual 3D scenes. In this context, the seminar focusses on data driven methods, which use motion capture data taken from human volunteers. The full range of relevant aspects of such approaches is offered to interested participants for overview presentations or for presentations on specialized topics.
This reaches from motion capturing technologies, to post-processing of motion capture data, semantic annotation of motion capture data as well as motion modelling and synthesis. Statistical approaches for motion modelling are of interest as well as more recent approaches which base on deep learning. For interested participants specialized presentations on background mathematics like for example principle component analysis or Gaussian processes are possible, too.
The seminar type is classic in the sense that registered participants will present the assigned topics, and discuss the strength and weaknesses of presented approaches. In addition, there will be two dedicated opponents for each presentation of an assigned topic. Successful participation in the seminar will be graded
The seminar will be held together with the DFKI.
Please refer to their webpage for further information.