An ideal-observer model of human sound localization

Reijniers, Jonas; Vanderelst, Dieter; Jin, Craig; Carlile, Simon; Peremans, Herbert

doi:10.1007/s00422-014-0588-4

Cited by 27 publications

(56 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Depending on the frequency content of the sound and the egocentric location of its source, each of these cues alone may be spatially ambiguous, leading to localization errors (Blauert, 1997; Shinn-Cunningham et al, 2000; King et al, 2001). In line with work on Bayesian integration of both multisensory (e.g., Ernst and Banks, 2002; Alais and Burr, 2004; Körding et al, 2007; Rowland et al, 2007; Targher et al, 2012; Hollensteiner et al, 2015; Zhou et al, 2018) and unisensory (e.g., Landy and Kojima, 2001; Knill and Saunders, 2003; Hillis et al, 2004; Watt et al, 2005; Sturz and Bodily, 2010) spatial cues, the localizing brain can be usefully compared with an ideal observer, integrating sound localization cues to resolve ambiguities by weighting each cue according its relative reliability (Reijniers et al, 2014).…”

Section: Introductionmentioning

confidence: 99%

Re-weighting of Sound Localization Cues by Audiovisual Training

Kumpik

Campbell

Schnupp

et al. 2019

Preprint

View full text Add to dashboard Cite

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest Author contribution statement DK, JS and AK conceived and designed the study; DK and CC carried out the experiments and analyzed the data; DK, JS and AK wrote the manuscript. Abstract Word count: 350Sound localization requires the integration in the brain of auditory spatial cues generated by interactions with the external ears, head and body. Perceptual learning studies have shown that the relative weighting of these cues can change in a contextdependent fashion if their relative reliability is altered. One factor that may influence this process is vision, which tends to dominate localization judgments when both modalities are present and induces a recalibration of auditory space if they become misaligned. It is not known, however, whether vision can alter the weighting of individual auditory localization cues. Using non-individualized head-related transfer functions, we measured changes in subjects' sound localization biases and binaural localization cue weights after ~55 minutes of training on an audiovisual spatial oddball task. Four different configurations of spatial congruence between visual and auditory cues (interaural time differences (ITDs) and frequency-dependent interaural level differences (interaural level spectra, ILS) were used. When visual cues were spatially congruent with both auditory spatial cues, we observed an improvement in sound localization, as shown by a reduction in the variance of subjects' localization biases, which was accompanied by an up-weighting of the more salient ILS cue. However, if the position of either one of the auditory cues was randomized during training, no overall improvement in sound localization occurred. Nevertheless, the spatial gain of whichever cue was matched with vision increased, with different effects observed on the gain for the randomized cue depending on whether ITDs or ILS were matched with vision. As a result, we observed a similar up-weighting in ILS when this cue alone was matched with vision, but no overall change in binaural cue weighting when ITDs corresponded to the visual cues and ILS were randomized. Consistently misaligning both cues with vision produced the ventriloquism aftereffect, i.e., a corresponding shift in auditory localization bias, without affecting the variability of the subjects' sound localization judgments, and no overall change in binaural cue weighting. These data show that visual contextual information can invoke a reweighting of auditory localization cues, although concomitant improvements in sound localization are only likely to accompany training with fully congruent audiovisual information. Contribution to the fieldFor the brain to construct a single unified percept of the world, it is necessary that information from our different senses is combined and integrated. This is often thought to be achieved in a statistically optimal fashion with different se...

show abstract

Section: Introductionmentioning

confidence: 99%

Re-weighting of Sound Localization Cues by Audiovisual Training

Kumpik

Campbell

Schnupp

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…The frequency range for the generic composite feature-based approach is selected empirically, where [0,4] kHz and [3,5] kHz are the phase and magnitude feature regions for the feature-based method, while the full-band signal is used for the correlation approach. During the comparison, the mean angular error is employed as a metric to assess the localization performance.…”

Section: Simulation Configurationmentioning

confidence: 99%

“…To adapt those devices for great spatial experiences, it is necessary to find the most valuable acoustic cues for human hearing system [1][2][3][4]. Additionally, testing and evaluating the quality of reproduced sound field and the hearing experience of those devices is another challenging problem [5][6][7]. The hearing tests with real human volunteers can be time-consuming and would be unfeasible for a huge amount of devices and testing scenarios.…”

Section: Introductionmentioning

confidence: 99%

Individualized Interaural Feature Learning and Personalized Binaural Localization Model

Talagala

Zhang

et al. 2019

Applied Sciences

View full text Add to dashboard Cite

The increasing importance of spatial audio technologies has demonstrated the need and importance of correctly adapting to the individual characteristics of the human auditory system, and illustrates the crucial need for humanoid localization systems for testing these technologies. To this end, this paper introduces a novel feature analysis and selection approach for binaural localization and builds a probabilistic localization mapping model, especially useful for the vertical dimension localization. The approach uses the mutual information as a metric to evaluate the most significant frequencies of the interaural phase difference and interaural level difference. Then, by using the random forest algorithm and embedding the mutual information as a feature selection criteria, the feature selection procedures are encoded with the training of the localization mapping. The trained mapping model is capable of using interaural features more efficiently, and, because of the multiple-tree-based model structure, the localization model shows robust performance to noise and interference. By integrating the direct path relative transfer function estimation, we propose to devise a novel localization approach that has improved performance in the presence of noise and reverberation. The proposed mapping model is compared with the state-of-the-art manifold learning procedure in different acoustical configurations, and a more accurate and robust output can be observed.

show abstract

“…There is a significant improvement in speech comprehension when distracters are in the opposite hemifield as opposed to the same hemifield (Hawley et al 1999). Although an opposite hemifield configuration would provide more localization information because of the position of the ears (e.g., Reijniers et al 2014), this advantage may be enhanced by the ascending auditory system: presentation in opposing hemifields would lead to preferential processing in the contralateral ICs. We would predict that there is a substantial, qualitative, difference between processing of stimuli at midline (where the signal would be equally represented in both ICs) and stimuli presented away from midline (where the signal would be represented primarily in one IC).…”

Section: Monaural and Binaural Processing Of Harmonic Complexesmentioning

confidence: 99%

A function for binaural integration in auditory grouping and segregation in the inferior colliculus

Nakamoto

Shackleton

Magezi

et al. 2015

Journal of Neurophysiology

View full text Add to dashboard Cite

Nakamoto KT, Shackleton TM, Magezi DA, Palmer AR. A function for binaural integration in auditory grouping and segregation in the inferior colliculus. J Neurophysiol 113: 1819 -1830, 2015. First published December 24, 2014 doi:10.1152/jn.00472.2014.-Responses of neurons to binaural, harmonic complex stimuli in urethaneanesthetized guinea pig inferior colliculus (IC) are reported. To assess the binaural integration of harmonicity cues for sound segregation and grouping, responses were measured to harmonic complexes with different fundamental frequencies presented to each ear. Simultaneously gated harmonic stimuli with fundamental frequencies of 125 Hz and 145 Hz were presented to the left and right ears, respectively, and recordings made from 96 neurons with characteristic frequencies Ͼ2 kHz in the central nucleus of the IC. Of these units, 70 responded continuously throughout the stimulus and were excited by the stimulus at the contralateral ear. The stimulus at the ipsilateral ear excited (EE: 14%; 10/70), inhibited (EI: 33%; 23/70), or had no significant effect (EO: 53%; 37/70), defined by the effect on firing rate. The neurons phase locked to the temporal envelope at each ear to varying degrees depending on signal level. Many of the cells (predominantly EO) were dominated by the response to the contralateral stimulus. Another group (predominantly EI) synchronized to the contralateral stimulus and were suppressed by the ipsilateral stimulus in a phasic manner. A third group synchronized to the stimuli at both ears (predominantly EE). Finally, a group only responded when the waveform peaks from each ear coincided. We conclude that these groups of neurons represent different "streams" of information but exhibit modifications of the response rather than encoding a feature of the stimulus, like pitch.

show abstract

An ideal-observer model of human sound localization

Cited by 27 publications

References 32 publications

Re-weighting of Sound Localization Cues by Audiovisual Training

Re-weighting of Sound Localization Cues by Audiovisual Training

Individualized Interaural Feature Learning and Personalized Binaural Localization Model

A function for binaural integration in auditory grouping and segregation in the inferior colliculus

Contact Info

Product

Resources

About