The natural human-human behavior can be regarded as the gold-standard for system performance and robustness (Huang and Mutlu, 2012) and thus, the analysis of natural behavior is required for developing a model of social interaction. Using either handcrafted models for distinguishing a person who aims at initiating an interaction from people who do not wish to interact and/or deriving models from lab data, did not work as intended in the real world (Michalowski et al., 2006 Bohus and Horvitz, 2009a). These signals could be very subtle, e.g., if a customer sits at the bar and decides to order another drink, s/he might not get up or move to another location. Thus, the system should not only detect the right signals, but also avoid false alarms. On the other hand, inviting customers to place an order if they had no intention to do so is annoying for those customers. Detecting customers who wish to order is crucial because failing to do so is fatal for the interaction as a whole. This is complicated by the fact that bars are often dimly-lit and noisy environments with multiple customers. In the bar scenario, one of the most difficult challenges is to distinguish between customers who are intending to place an order and those who are not. We conclude that (a) these two easily recognizable actions are sufficient for recognizing the intention of customers to initiate a service interaction, but other actions such as gestures and speech were not necessary, and (b) the use of reaction time experiments using natural materials is feasible and provides ecologically valid results.įor enabling users to interact intuitively with a robotic agent, the robot system has to be able to identify and to respond to social signals appropriately. Finally, a signal detection analysis revealed that ignoring a potential order is deemed worse than erroneously inviting customers to order. The participants also showed a strong agreement about when these cues occurred in the videos. Both signals were necessary and, when occurring together, sufficient. The results revealed that bar staff responded to a set of two non-verbal signals: first, customers position themselves directly at the bar counter and, secondly, they look at a member of staff. Two experiments using snapshots and short video sequences then tested the validity of these hypothesized candidate signals. These recordings were used to generate initial hypotheses about the signals customers produce when bidding for the attention of bar staff. In order to study which signals customers use to initiate a service interaction in a bar, we recorded real-life customer-staff interactions in several German bars. Thus, a bartending robot has to be able to distinguish between customers intending to order, chatting with friends or just passing by. This detection is particularly challenging in a noisy environment with multiple customers. Detecting whether a customer would like to order is essential for the service encounter to succeed. Enabling a bartending robot to serve customers is particularly challenging as the system has to recognize the social signals produced by customers and respond appropriately. Recognizing the intention of others is important in all social interactions, especially in the service domain.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |