Body Pose Predictions in Triadic Social Interactions

Human beings are social animals in that they need to socialize with each other to build companionship and thrive alongside other humans. One of the primary characteristics of social interactions is the signals used by people to communicate their thoughts effectively. These include gesturing with their hands, moving around etc.. AI agents or algorithms interacting with humans which we refer to as Social artificial intelligence must learn to interpret and predict these signals in order to use them to interact with other humans successfully. Data-driven approaches have helped make remarkable strides in many artificial intelligence tasks and could similarly help machines learn the body gestures of interacting individuals. We define a framework for predicting these gestures in a triadic social interactions scenario where the humans play a game of haggling and two sellers try to sell their products to a buyer.

Author dc:creator

• Girdhar, Rishab

• Human-Robot Interaction
• Pose Prediction
• Social Artificial Intelligence

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/11299/223089

OAI identifier oai:identifier

oai:conservancy.umn.edu:11299/223089