Original article was published on Communications of the ACM – Artificial Intelligence
May 20, 2020
University of Maryland (UMD) researchers have developed an algorithm to enable a small four-wheeled robot to perform real-time gait analysis in order to guess a walker’s emotional state.
The researchers collected information from humans who observed each other while walking, inquired about each other’s emotional state, and correlated that information with data about each subject’s gait.
The team then used videos of subjects walking, combined with connections between skeletal gaits and emotions associated with those walkers, to train the ProxEmo algorithm.
UMD’s Aniket Bera suggested robots equipped with the algorithm could intervene, or move out of the way, after determining a person’s emotional state by reading their gait.
View Full Article
Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA
No entries found