Auditory Perspective Taking
Abstract
Auditory perspective taking is imagining being in another's place, and predicting what they are able to hear and how it will affect their general comprehension. From this knowledge of another's auditory perspective, a conversational partner can then adapt his or her auditory output to overcome a variety of environmental challenges and insure that what is said is intelligible. In collaborative activities, perspective taking skills greatly facilitate the effort participants must make in communicating with each other. By focusing on perspective taking in the auditory domain, one seeks to allow a robot to use its knowledge of the environment, both a priori and sensed, to predict what its human counterpart can hear and effectively understand. Equipped with this knowledge, the robot can change its auditory presentation behavior accordingly. In the case of the helicopter flying overhead, for instance, the robot should be able to predict the inability of its addressee to adequately hear what it is saying and either try to talk louder or simply pause until noise levels return to normal. Either option is something that is easily implemented, and which could go far towards improving the quality of human-robot interaction. In general, a robot capable of auditory perspective taking could pay attention to a number of different factors that might affect human robot communication: masking noises, interruptions/distractions, changes in operator or robot position, and individual differences between different operators. As people can adapt to each of these naturally, a robot speech interface unable to similarly adapt is likely to prove frustrating to a human partner. In this poster presentation, the authors explore the concept of auditory perspective taking using an audio-visual robot speech interface.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2006
- Accession Number
- ADA462844
Entities
People
- Derek Brock
- Eric Martinson
Organizations
- United States Naval Research Laboratory