Everyday Activity in Science and Engineering
As part of the EASE collaborative research center we are investigating how humans perform everyday activities. We are collecting a dataset with various different sensors of subjects tasked with setting a table for various meal types and conditions. The dataset will be made publicly available. The data is then semantically annotated on different levels of granularity and abstraction in order to generate an action plan for a task. These symbolic representations of the everyday tasks can be used by a humanoid robots to enable or enhance its performance of those same actions.
Sensors used in the current experimental setup include a OptiTrack optical motion capture, stationary and head
mounted cameras, microphones for scene and voice audio, and a variety of worn sensors to measure bio-physiological signals produced by the human body during activity. These include accelerometers, electromyography (EMG) and electrodermal activity (EDA) sensors, a g.Tec Nautilus mobile electroencephalography (EEG) cap, and a Pupil Labs mobile eye-tracker.
Mehr Informationen: https://ease-crc.org/