Modeling Social Behaviors Researchers: Faisal Khan Human social behavior is complex, dynamic, and shaped by context, culture, and social and cognitive processes. Understanding social behavior computationally requires methods and models that can capture this complexity. This project focuses on using stochastic methods for modeling social cues. These models are used as bases for designing social norms and behaviors for humanlike agents and robots.

Brain-Robot Interfaces Researchers: Jonathan Mumm Neural signals such as electrical activity captured through Electroencephalography (EEG) and oxygenated hemoglobin captured through functional near-infrared imaging (fNIR) allow researchers to measure changes in cognitive, attentional, and affective states. There is growing interest in using these signals as input for computing applications as predictors of mental states (e.g., at rest, working) or for controlling pointing devices solely through neural activity. We are interested in exploring how neural signals could be used to improve communication with robots in collaborative tasks.

Human-Robot Proxemics Researchers: Jonathan Mumm, Philip Zhao Physical embodiment is a unique quality that humanlike robots possess and that affect how people perceive and interact with them. The presence of this quality leads people to expect the appropriate proxemic behavior of robots. Our previous work has shown that robots can be detrimental to people’s work practices and social interactions. The goal of this project is to gain a better understanding of and to develop design specifications for appropriate proxemic behavior for humanlike robots.

Studying Rapport and Trust in Human-Agent Interaction Researchers: Jonathan Mumm, Philip Zhao Humanlike agents and robots are envisioned as educational assistants, personal trainers, life coaches. In this project, we examine how information on a person’s performance should be presented by an embodied conversational agent (ECA) to establish and maintain support trust, rapport, and cooperation between the person and the agent.

Designing Public Social Behavior for Robots Researchers: Irene Shei Every day in passing we, as humans, are constantly signaling to others through to communicate our availability and interest. Walking along the street, we use our body language, facial expression, and even gaze to let others know that we’re busy or to show that we’d like to stop and say hello. As a result, we’ve become experts at assimilating and interpreting this information at a glance and using that analysis to formulate our own actions and reactions. This experiment investigates whether or not the same behavioral reactions hold true when our partners are robots to improve human-robot social interaction.

Investigating Basic Social Processes Using Agents and Robots Researchers: Megan DiVall When people interact with agents and robots, they tend to use communicative processes and follow social norms that are similar to those we observe in human-human interaction. This tendency allows us to use agents and robots to conduct basic research in social cognition and interpersonal communication. In this project, we study phenomena from human social communication by creating situations in which people interact with agents or robots and by testing hypotheses from social and cognitive psychology and communication research. We currently focus on deception and the cues that people use to determine that their partner might be dissembling.

Designing Effective Educational Agents Researchers: Andrego Halim Research in education suggest that social cues, particularly embodied cues such as gaze and gestures, play an important role in learning and attention in the classroom. In this project, we seek to obtain a computational understanding of such cues and explore how educational agents might effectively use them to improve attention and learning.