Kathrin Janowski

Empathy Simulation

Duration:	2013 - 2016
Involved Scientists:	M.Sc. Kathrin Janowski, Prof. Dr. Hannes Ritschel, Dr. Florian Lingenfelser, Dr. Johannes Wagner, M.Sc. Markus Häring, Prof. Dr. Birgit Lugrin

Description

The simplest form of empathy is to mirror the emotions that are observed in the conversation partner. For humans, this usually happens subconsciously.

Our prototype used the Social Signal Interpretation Framework (SSI) to analyze the user's tone of voice and facial expressions. The recognized emotions were then mapped to numerical values along the axes of "Pleasure" and "Arousal".

We used two RoboKind R-50 robots from the Texas-based company Hanson Robotics."Alice" and "Zeno" each had a human-like body and additional motors under synthetic skin that allowed for adjusting their facial expressions. The application that controlled their behavior continuously matched their expressions and head pose to the emotional state detected by the SSI pipeline.

Furthermore, several lists of sentences were provided that the robots could use to comment on the situation when the user paused in speaking. The exact timing of the robots' speech depended on different factors, such as the intensity of the current emotional state or the time elapsed since its last comment.