Sound Localization and Separation in 3D Space Using a Single Microphone with a Metamaterial Enclosure

Abstract

Conventional approaches to sound localization and separation are based on microphone arrays in artificial systems. Inspired by the selective perception of the human auditory system, a multisource listening system which can separate simultaneous overlapping sounds and localize the sound sources in 3D space, using only a single microphone with a metamaterial enclosure is designed. The enclosure modifies the frequency response of the microphone in a direction‐dependent manner by giving each direction a characteristic signature. Thus, the information about the location and the audio content of sound sources can be experimentally reconstructed from the modulated mixed signals using a compressive sensing algorithm. Due to the low computational complexity of the proposed reconstruction algorithm, the designed system can also be applied in source identification and tracking. The effectiveness of the system in multiple real‐life scenarios is evaluated through multiple random listening tests. The proposed metamaterial‐based single‐sensor listening system opens a new way of sound localization and separation, which can be applied to intelligent scene monitoring and robot audition.

LSTM‐based japanese speaker identification using an omnidirectional camera and voice information

“…However, the usefulness of the proposed approach has not been shown for a case with two subjects speaking simultaneously. In such cases of simultaneous speech by multiple people, the individual voices should be separated using the voice separation method [32][33][34], and the proposed method then applied. These are potential avenues for further research.…”

A wave-confining metasphere beamforming acoustic sensor for superior human-machine voice interaction

“…They represent an enhanced acoustic sensing strategy by pressure amplification using different passive metamaterial structures. On the other hand, a single regular sensor inside spatially designed acoustic metamaterials has been shown for the separation and localization of sound waves ( 25 , 26 ) and elastic vibrations ( 27 ) without using arrayed acoustic sensors. In these systems, either a metamaterial enclosure or a resonant coupling network encodes the response of the sensor in a direction-dependent manner, followed by a demodulation algorithm to reconstruct the source locations.…”

2D phononic-crystal Luneburg lens for all-angle underwater sound localization

“…The traditional approach for localizing a sound source is hydrophone arrays according to the time difference of receiving signal [28,29]. To overcome that the accuracy of sound localization is limited by the number of the microphones and the array scale [30], researchers attempt to use the metamaterial to develop new sound localization techniques for more application scenarios [31,32]. In this work, we designed square-lattice and annular-lattice Luneburg lenses based on GRIN phononic crystals whose unit cell is made of PLA and air inclusion in a water environment, and we proposed a novel application of GRIN phononic crystals to sound direction identification and sound localization covering 360°.…”