Recovering Visual Human's Attention in Human-Robot Interaction for a Promoter Service Robot

The use of service robots is an alternative for prod-ucts and services promotion. In commercial areas an effective human-robot interaction is essential. However, communicating information effectively and thus maximizing the likelihood of a product or service's acceptance by the potential consumer remains challenging. In this work, we propose an architecture that integrates multiple modules, including visual gaze esti-mation, head tracking, human detection, position estimation, and attention recovery behaviors, to recover human attention during potential scenarios such as promotional tasks. Attention is presented as a decay function on different parameters, such as visual gaze and head position. When the robot detects a shutdown at the attention level, it executes a behavior to recover it. The proposal details our architecture's design and the modules' theoretical implications, highlighting the potential for improved service-robot interaction.