DATA_PAUSCH2022 - - Results from [1]

Usage

data = data_pausch2022(type_stf,type_mod,fig,plot)

Input parameters

type_stf

flag that may be used to choose between one of the following:

hrtf load individual HRTF datasets
hartf_front load invidual front HARTF datasets
hartf_rear load invidual rear HARTF datasets

Output parameters

data

data struct

(if type_stf=={'hrtf','hartf_front','hartf_rear'}) Individual spatial transfer functions, with fields id, the participant ID and sofa, the sofa file
(if weights=='weights') Polynomial regression weights (polynomial degree of P=4), to be applied on a subset of M=5 individual features
if type_mod=={'kuhn','woodworth','woodworth_ext'}: vector of (M*P+1) x 1 weights, if type_mod=='pausch': matrix of (M*P+1) x 4 weights
(if features=='features') Individual features as published in [1]

itd

itd struct (if type_stf=={'hrtf','hartf_front','hartf_rear'}) Estimated ITDs for directions in the horizontal plane

itd_hor: ITDs in s [double]
itd_max: maximum ITD in s [s]
itd_arg_max_idx: index of the argument of the the maximum ITD [double]
itd_arg_max_phi: argument of the the maximum ITD in deg [double]
itd_hor_mean: direction-dependent mean ITDs across participants in s [double]
itd_hor_std: direction-dependent standard deviation of ITDs across participants in s [double]
bp_fc_low: lower cut-off frequency (Hz) of the bandpass filter applied before estimating the ITDs [double]
bp_fc_high: upper cut-off frequency (Hz) of the bandpass filter applied before estimating the ITDs [double]

Description

The type_mod flag may be used to select the ITD model for the ITD predictions (only required if weights=='weights'):

`'pausch'`	hybrid ITD model by Pausch et al. [1] (default)
`'kuhn'`	analytic ITD model by Kuhn [2]
`'woodworth'`	analytic ITD model by Woodworth [3]
`'woodworth_ext'`	analytic ITD model by Woodworth [3] extended by Aaronson and Hartmann [4]

The weights flag may be one of the following:

`'no_weights'`	do not load polynomial regression weights (default)
`'weights'`	load polynomial regression weights

The features flag may be one of the following:

`'no_features'`	do not the individual features (default)
`'features'`	load the individual features

Additional key/value pairs include:

`'bp_fc_low'`	lower cut-off frequency (Hz) of the bandpass filter applied before estimating the ITDs (default: []) [double]
`'bp_fc_high'`	upper cut-off frequency (Hz) of the bandpass filter applied before estimating the ITDs (default: []) [double]

Requirements:

SOFA API v0.4.3 or higher from http://sourceforge.net/projects/sofacoustics for Matlab (in e.g. thirdparty/SOFA)
Data in hrtf/pausch2022 and auxdata/pausch2022 (downloaded on the fly)

References:

F. Pausch, S. Doma, and J. Fels. Hybrid multi-harmonic model for the prediction of interaural time differences in individual behind-the-ear hearing-aid-related transfer functions. Acta Acust., 2022. Under review.