Binaural Sound Source Localization Research Articles

Improved binaural sound localization and tracking for unknown time-varying number of speakers

A method based on the generalized cross-correlation (GCC) method weighted by the phase transform (PHAT) has been developed for binaural sound source localization (SSL) and tracking of multiple sound sources. Accurate binaural audition is important for applying inexpensive and widely applicable auditory capabilities to robots and systems. Conventional SSL based on the GCC-PHAT method is degraded by low resolution of the time difference of arrival estimation, by the interference created when the sound waves arrive at a microphone from two directions around the robot head, and by impaired performance when there are multiple speakers. The low-resolution problem is solved by using a maximum-likelihood-based SSL method in the frequency domain. The multipath interference problem is avoided by incorporating a new time delay factor into the GCC-PHAT method with assuming a spherical robot head. The performance when there are multiple speakers was improved by using a multisource speech tracking method consisting of voice activity detection (VAD) and K-means clustering. The standard K-means clustering algorithm was extended to enable tracking of an unknown time-varying number of speakers by adding two additional steps that increase the number of clusters automatically and eliminate clusters containing incorrect direction estimations. Experiments conducted on the SIG-2 humanoid robot show that this method outperforms the conventional SSL method; it reduces localization errors by 18.1° on average and by over 37° in the side directions. It also tracks multiple speakers in real time with tracking errors below 4.35°.

Cepstrum Prefiltering for Binaural Source Localization in Reverberant Environments

Binaural sound source localization can be performed by imitation of the fundamental mechanisms of the human auditory system, which is based on the integrated effects of ear, pinnae, head and torso. In particular, two physical cues can be exploited, i.e. the Interaural Time Difference (ITD) and the Interaural Level Difference (ILD). It is known that joint use of ITD and ILD provides good source azimuth estimations. In many practical situations binaural localization has to be performed in closed environments, where the presence of reverberation degrades the performance of available position estimators. In this paper a possible solution to this difficult problem is introduced. The proposed solution is based on proper use of cepstral prefiltering prior to source localization by ITD and ILD. It is shown that cepstrum can help in reducing the effects of reverberation, thus yielding better location estimates.

A biologically inspired spiking neural network model of the auditory midbrain for sound source localisation

This paper proposes a spiking neural network (SNN) of the mammalian subcortical auditory pathway to achieve binaural sound source localisation. The network is inspired by neurophysiological studies on the organisation of binaural processing in the medial superior olive (MSO), lateral superior olive (LSO) and the inferior colliculus (IC) to achieve a sharp azimuthal localisation of a sound source over a wide frequency range. Three groups of artificial neurons are constructed to represent the neurons in the MSO, LSO and IC that are sensitive to interaural time difference (ITD), interaural level difference (ILD) and azimuth angle ( θ ) , respectively. The neurons in each group are tonotopically arranged to take into account the frequency organisation of the auditory pathway. To reflect the biological organisation, only ITD information extracted by the MSO is used for localisation of low frequency ( < 1 kHz ) sounds; for sound frequencies between 1 and 4 kHz the model also uses ILD information extracted by the LSO. This information is combined in the IC model where we assume that the strengths of the inputs from the MSO and LSO are proportional to the conditional probability of P ( θ | ITD ) or P ( θ | ILD ) calculated based on the Bayes theorem. The experimental results show that the addition of ILD information significantly increases sound localisation performance at frequencies above 1 kHz. Our model can be used to test different paradigms for sound localisation in the mammalian brain, and demonstrates a potential practical application of sound localisation for robots.

Binaural Active Audition for Humanoid Robots to Localise Speech over Entire Azimuth Range

We applied motion theory to robot audition to improve the inadequate performance. Motions are critical for overcoming the ambiguity and sparseness of information obtained by two microphones. To realise this, we first designed a sound source localisation system integrated with cross-power spectrum phase (CSP) analysis and an EM algorithm. The CSP of sound signals obtained with only two microphones was used to localise the sound source without having to measure impulse response data. The expectation-maximisation (EM) algorithm helped the system to cope with several moving sound sources and reduce localisation errors. We then proposed a way of constructing a database for moving sounds to evaluate binaural sound source localisation. We evaluated our sound localisation method using artificial moving sounds and confirmed that it could effectively localise moving sounds slower than 1.125 rad/s. Consequently, we solved the problem of distinguishing whether sounds were coming from the front or rear by rotating and/or tipping the robot's head that was equipped with only two microphones. Our system was applied to a humanoid robot called SIG2, and we confirmed its ability to localise sounds over the entire azimuth range as the success rates for sound localisation in the front and rear areas were 97.6% and 75.6% respectively.

A real-time model of binaural sound source localization resolving spatial ambiguities

The human binaural hearing provides a precise internal image of external sound source locations. Although the determination of the azimuthal position of a sound source is primarily based on the evaluation of interaural time and intensity differences, the psychoacoustic processing of these cues to form an ‘‘auditory spatial image’’ is yet unclear. Especially for all angles of incidence on the ‘‘cone of confusion,’’ interaural differences are equal and therefore provide an ambiguous representation of the source location. However, neurophysiological studies with barn owls show that by integration of directional information across frequency bands an unambiguous internal representation of sound source locations can be obtained. These findings form a model of binaural localization which was adapted to the system characteristics and psychoacoustic parameters of human hearing. The model was implemented in real-time on a multiprocessor setup. First results with the localization of sounds under anechoic and normal room conditions indicate that incidence directions from ahead, above and from the back can be distinguished. A possible use of a preprocessing stage in a directional hearing aid is discussed.