This study assessed the feasibility of video-based tracking of eye gaze and facial components during robot assisted gait walking. Feature-based tracking of user’s face, eyes, nose, mouth and pupils has been implemented and applied to video recordings of four healthy controls during four minutes long walking on Hocoma’s Lokomat. Four different visual feedback modalities were tested (no visual feedback, predefined gaze targets and 3D virtual environment from the first and third person perspective). On average, user’s face was detected in ~95% of video frames with provided visual feedback. Eyes, mouth and nose were detected in ~99% of detected faces. Pupils were detected in ~100% of detected eyes.
On Feasibility of Gaze and Facial Component Tracking during Robot Assisted Walking
Biosystems & Biorobotics
Converging Clinical and Engineering Research on Neurorehabilitation ; Kapitel : 48 ; 305-309
2013-01-01
5 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Reflex assisted walking for a hexapod robot
IEEE | 2016
|Walking robot, method for controlling walking robot, and walking robot system
Europäisches Patentamt | 2023
|