Dynamically reconfigurable microphone arrays

Abstract

Robotic sound localization has traditionally been restricted to either on-robot microphone arrays or embedded microphones in aware environments, each of which have limitations due to their static configurations. This work overcomes the static configuration problems by using visual localization to track multiple wireless microphones in the environment with enough accuracy to combine their auditory streams in a traditional localization algorithm. In this manner, microphones can move or be moved about the environment, and still be combined with existing on-robot microphones to extend array baselines and effective listening ranges without having to re-measure inter-microphone distances.

On analyzing user location discovery methods in smart homes: A taxonomy and survey

“…Sound plays an important role in locating a speaker in a wide target range, and thus sound ULD systems with a microphone array have been developed in various forms. However, a robot usually has space restrictions when mounting a microphone array on its head; this increases the time delay of arrival (TDOA) error and decreases the resolution of sound ULD [45]. Vision, on the contrary, has directional limitation, i.e., vision can only detect users inside the visible range.…”

Optimizing a reconfigurable robotic microphone array

“…Nakadai et al [7] have integrated robotic sound localization results with room mounted arrays also for tracking, but without synchronizing streams between the two. Martinson and Fransen [8] synchronized streams between robot mounted microphone arrays and scattered wireless microphones, but did not address how to improve an initially poor configuration.…”

“…A dynamically reconfigurable microphone array [8] is an array that can change its physical properties, including number and positioning of microphones, in response to external stimuli. In this work, such an array is constructed with three Pioneer3-AT robots (Figure 1, top).…”

“…These same fiducials were also used in combination with robot odometry to localize individual robot team members. The localization error for such a system is generally less than 1cm, and is presented in greater detail in [8].…”