Fostering positive team behaviors in human-machine teams through emotion processing: Adapting to the operator s state
Abstract
The proposed project is a collaborative effort between RMIT University and AFRL. Based on RMIT’s speech research in which the emotional dynamics of a dyad can be estimated computationally, the following project aims to monitor preferred teammate speech dynamics and emulate this pattern in an autonomous partner. This would allow for the investigation of the following research questions.Question 1: Without providing trustworthiness information, would a subject assume trustworthiness based on the dynamics of communication? For example, could a subject identify variances in trustworthiness based on communication style, e.g. variance in ability, benevolence and integrity?Question 2: Does trust promoting interaction from a robotic partner result in trust or distrust due to a violation of expectations; robots are not empathetic or supportive?Question 3: How does stress impact the importance of optimal dyadic communication? Under stressful conditions with novel decision-making scenarios, emotive and optimal partner communication may reduce experienced negative emotion. This change in state would free resources for decision-making and would remove bias in processing.Question 4: Can a subjective assessment of trustworthiness be supported by objective measures derived from emotive modelling of audiovisual data (speech and facial images)?Question 5: To what extent is the modelling of emotive interactions in conversations correlated with the subjective perception of trustworthiness in human–human and human–machine teams?
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- May 02, 2017
- Source ID
- FA23861710095
Entities
People
- Margaret Lech
Organizations
- Air Force Office of Scientific Research
- RMIT University
- United States Air Force