Forschungsthemen
[BA] A Metamodel for Describing Human Activities of Daily Living
Human movement is an essential component of our daily lives. It enables us to survive through reflexes, to interact with objects through targeted movements, to explore our surroundings by locomotion, and to communicate with other people by face expressions and gestures, also with the help of technical devices. It is therefore not surprising that human movement is a subject of research in various disciplines involving biomechanics, ergomonics, psychology, rehabilitation science, robotics, computergraphics and human-machine interaction. Different taxonomies were developed that characterize aspects of human movements such as structural or temporal kinematic features and motion style. However, they are often applied in a discipline-specific manner and do not provide a format or modelling approach that can be shared with and reused by the research community. Thus, despite this great interest in human movement, there is still no description language available that can characterize human movements in a standarized but flexible manner.
That is why we are seeking for the development of a first metamodel to describe human movements. For this approach we are focusing on activities of daily living that involve object manipulation and body transfer (i.e. no face expressions). On the long run, a metamodel of human movement can open up new possibilities at all processing steps of dealing with motion data, e.g., in the design of data sets by describing and planning movements before actually capturing them, in annotating movements with standardized labels that are provided by the metamodel, in automatically deriving features from the description of movements, and in generating synthetic movements from standardized sets of labels.
Individual Tasks
Betreuer: Johannes Mey and Loreen Pogrzeba