What we see is what we do: A peripheral vision based HMM framework for gaze guided human action recognition
FELIX WANG, Thomas Kreiner · Figshare · 2022
The dataset contains gaze data from a medical device handling task, which was used in a HMM based human action recognition (HAR) framework. It contains 43 data sequences from 20 participants, where 2 participants where expert in the observed task (Exp1 and Exp2). The data is given for each step of the HAR framework, as well as with added gaussian noise.