Abstract
The mammalian auditory system extracts features from the acoustic environment based on the responses of spatially distributed sets of neurons in the subcortical and cortical auditory structures. The characteristic responses of these neurons (linearly approximated by their spectro-temporal receptive fields, or STRFs) suggest that auditory representations are formed, as early as in the inferior colliculi, on the basis of a time, frequency, rate (temporal modulations) and scale (spectral modulations) analysis of sound. However, how these four dimensions are integrated and processed in subsequent neural networks remains unclear. In this work, we present a new methodology to generate computational insights into the functional organization of such processes. We first propose a systematic framework to explore more than a hundred different computational strategies proposed in the literature to process the output of a generic STRF model. We then evaluate these strategies on their ability to compute perceptual distances between pairs of environmental sounds. Finally, we conduct a meta-analysis of the dataset of all these algorithms' accuracies to examine whether certain combinations of dimensions and certain ways to treat such dimensions are, on the whole, more computationally effective than others. We present an application of this methodology to a dataset of ten environmental sound categories, in which the analysis reveals that (1) models are most effective when they organize STRF data into frequency groupings—which is consistent with the known tonotopic organization of receptive fields in auditory structures -, and that (2) models that treat STRF data as time series are no more effective than models that rely only on summary statistics along time—which corroborates recent experimental evidence on texture discrimination by summary statistics.
Highlights
The mammalian auditory system extracts features from the acoustic environment based on the responses of spatially distributed sets of neurons in the inferior colliculi (IC), auditory thalami and primary auditory cortices (A1)
These two computational trends are in interesting accordance with the tonotopical organization of spectro-temporal receptive field (STRF) in central auditory structures (Eggermont, 2010; Ress and Chandrasekaran, 2013) as well as recent findings on texture discrimination by summary statistics (McDermott et al, 2013; Nelken and de Cheveigné, 2013)
This suggests that metaanalysis over a space of computational models can generate insights that would otherwise be overlooked in a field where current results are scattered, having been developed with different analytical models, fitting methods and datasets
Summary
The mammalian auditory system extracts features from the acoustic environment based on the responses of spatially distributed sets of neurons in the inferior colliculi (IC), auditory thalami and primary auditory cortices (A1). The authors found that their model approximates psychoacoustical dissimilarity judgements made by humans between pairs of sounds to near-perfect accuracy, and better so than alternative models based on simpler spectrogram representation Such computational studies (see Fishbach et al, 2003) provide proofs that a given combination of dimensions (e.g., frequency/rate/scale for Patil et al, 2012; frequency/rate for Fishbach et al, 2003), and a given processing applied on it, is sufficient to give good performance; they do not, answer the more general questions of what combination of dimensions is optimal for a task, in what order these dimensions are to be integrated, or whether certain dimensions are best summarized rather than treated as an orderly sequence. We propose a systematic pattern-recognition framework to, first, design more than a hundred different computational strategies to process the output of a generic STRF model; second, we evaluate each of these algorithms on their ability to compute acoustic dissimilarities between pairs of sounds; third, we conduct a meta-analysis of the dataset of these many algorithms’ accuracies to examine whether certain combinations of dimensions and certain ways to treat such dimensions are more computationally effective than others
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
