Hierarchical model of pitch perception using Top-Down Modulation.

From Emcap

Contents

Understanding Pitch Perception as a Hierarchical Process with Top-Down Modulation

by Balaguer-Ballester E., Clark, N., Coath, M., Krumholz, K. and Denham

University of Plymouth, University of Nottingham

emili.balaguer@uv.es

emili.balaguer-ballester@plymouth.ac.uk

10-sept-2008

The software is currently unavailable for public download.

What is the Module?

This study presents a highly novel approach to computer modelling and understanding pitch perception. It provides a unitary account of the multiple time scales observed in temporal integration and temporal resolution in pitch perception. The model is evaluated using a wide range of perceptual studies and a neurophysiological experiment, which have not been previously accounted for by a single model. In contrast to other approaches, the current model contains a hierarchy of integration stages and uses feedback modulation to adapt the effective time scales of these integration stages to changes in the input stimulus. Therefore, the model could represent a form of generative approach and shows how efferent connections from central to subcortical areas may be critical for understanding the temporal dynamics of pitch processing in particular, and possibly auditory processing in general.

What are the inputs, outputs and parameters?

Parameters may be sustained to continuous improvements. Please see 'How to use' section for a comprehensive description of current version.

System requirements

Matlab, any recent version, no toolboxes and DSAM (see below).

How to Cite

NO CITATION YET BECAUSE MANUSCRIPT IS UNDER REVISION

Balaguer-Ballester E., Clark, N., Coath, M., Krumholz, K. and Denham, S. L., 2008. Understanding Pitch Perception as a Hierarchical Process with Top-Down Modulation.

Detailed information about the model can be found in the corresponding WP6 report, only accessible to reviewers.

How to use it

Aug 23, 2008. Version 3.0b

Emili Balaguer-Ballester

IMPORTANT: THIS SOFTWARE IS CURRENTLY CONFIDENTIAL BECAUSE THE MANUSCRIPT IN WHICH IS BASED IS STILL UNDER REVIEW. PLEASE DO NOT CIRCULATE. IT IS ALSO SUBJECT TO CONTINUOUS IMPROVEMENTS AND MIGTH CONTAIN SOME BUGS. UPDATES WILL BE AVAILABLE INMEDIATELY IF MANUSCRIPT ACCEPTED (see web adress below).

Work funded by EMCAP project (Emergence Cognition Through Active Perception, 2005-2008) http://emcap.iua.upf.es

Model described in the manuscrit 'Understanding Pitch Perception as a Hierarchical Procces with Top-Down Modulation. Not yet published.

This is a DSAM (Development System for Auditory Modelling) version. For a MAP (Matlab Auditory Peripheral) version please contact me (see below)

Note please that this is an implementation only for research and does not contain further refinement. It is quite slow. This is intentional, in order to make it clearer for the reader. For a faster implementation, please contact me.


1) Uninstall any previous AMS version, if any

2) Install AMS118v1.exe

http://www.pdn.cam.ac.uk/groups/dsam/dloads_main.htm

3) Extract the contents of this file (gpm3_0b.zip) into the root directory C:/gpm3_0b (not anywere else)

Important: All of the compressed files have to be located into the folder 'C:/gpm3_0b'.

If not, the results obtained later can be wrong, because can be a conflict with other installed applications.

4) Open Matlab, and add the path to this new directory ('File'-'Set Path'...). Browse in Matlab to this new directory.

5) Few demonstrative stimuli are located in the same compressed file, and can be tried by typing

'c:>gpm3_0b('RandomSequence.wav');'

etc.

For any question or giving feedback please contact corresponding author

Emili Balaguer-Ballester

Currently in:

emili.balaguer-ballester@plymouth.ac.uk

emili.balaguer@uv.es

For future updates (only if is accepted) and other software: http://www.pion.ac.uk/emili/index.htm If the page do not exists, please 'google' my name and surnames and you will find me easily.

Thanks very much.