NAME

gennap - generate neural activity pattern

CONTENTS

Synopsis
Description
Options
Neural Transduction
Examples
References
Files
See Also
Copyright
Acknowledgements

gennap [ option=value | -option ] [ filename ]

The gennap module of the AIM software converts an input wave into aneural activity pattern (NAP), which is AIM’s representation of thefiring pattern in the auditory nerve at about the level of thecochlear nucleus. Gennap begins by calculating the basilar membranemotion (BMM) associated with the input wave using the genbmm module,and then it applies several additional transforms that we know occurin some form during the neural transduction process. AIM provides twoalternative methods for generating the NAP, a two-dimensional adaptivethresholding mechanism (Holdsworth and Patterson, 1993), and an arrayof inner haircell simulators based on (Meddis et al., 1990; Giguereand Woodland, 1994). In the functional version of AIM, the adaptivethresholding mechanism applies rectification, log compression,adaptation in time, and suppression across frequency; its purpose isto stabilise the level of the membrane activity with compression andthen sharpen the features that appear in the compressed membranemotion. Together, the gammatone filterbank and adaptive thresholdingform a ’functional’ cochlea simulation. In the physiological versionof AIM, the transmission-line filterbank applies level-dependantcompression and the Meddis haircell bank applies compression andadaptation. The haircells are not coupled and so there is nofrequency sharpening in this module. Together, the transmission-linefilterbank and the Meddis module form a ’physiological’ cochleasimulation.

The options for gennap are grouped according to the functions theycontrol. The adaptive thresholding options are identified by thecommon suffix _at; the Meddis module options are identified by thecommon suffix _med. These two groups of options are the subject ofthis manual entry, together with two additional options that specifywhether rectification and compression are required beforetransduction. There is also an option to specify the choice oftransduction module.

The adaptive thresholding process in the functional version of AIMbegins with rectification and compression of the BMM. The defaultform of compression is logarithmic; it has the advantage oftransforming the exponential envelope of the ringing response of thegammatone filter into a linear decay with time. There is evidence,however, that auditory compression may be better represented by powercompression with an exponent in the range of 0.5. It is alsoadvisable in some cases to insert power compression before the Meddishaircell when driving it with a gammatone filter. For a discussion ofthese issues, see docs/aimMeddisHewitt. To accommodate the assemblyof different configurations of AIM, the rectification and compressionoptions are presented separately in the options list before the neuraltransduction section.

NOTE: When using the physiological version of AIM with thetransmission-line filterbank and the Meddis haircell bank, setcompress=off, as compression is an integral part of the feedback loopin the transmission-line filterbank module.

Neural Transduction

The neural transduction is performed either by two-dimensionaladaptive thresholding or an array of Meddis haircells. The choice iscontrolled by the option ’transduction’.

transduction	The transduction functionSwitch. Choices: at, med, off. Default value: at

If adaptive thresholding is specified (at), the options with suffix_at below apply; if the Meddis module is specified (med), the optionswith suffix _med below apply. If off is specified, no transductionfunction is applied. The default is at.

    II TWO-DIMENSIONAL ADAPTIVE THRESHOLDING: _at

The adaptive thresholding mechanism is a functional model of neuralencoding (Holdsworth, 1990; Patterson and Holdsworth, 1996). Itspurpose is to enhance the contrast of the larger features that appearin the surface of the BMM and reduce those aspects of therepresentation which are just a direct consequence of the filteringand compression processes. The process begins with rectification andcompression of the BMM. The tail of the envelope of the impulseresponse of the gammatone filter is exponential. As a result,logarithmic compression is used, since this makes the filter decayfunction linear in NAP coordinates. Following compression, adaptationis applied in time and suppression is applied across frequency(Holdsworth and Patterson, 1993, Patterson, 1994a).

Briefly, an adaptive threshold value is maintained for each channeland updated at the sampling rate. The new value is the largest of a)the previous value reduced by a fast-acting temporal decay factor(t1recovery_at), b) the previous value reduced by a longer-termtemporal decay factor (t2recovery_at), c) the adapted level in thechannel immediately above, reduced by a frequency spread factor(frecovery_at), d) the adapted level in the channel immediately below,reduced by the same frequency spread factor, or e) a floor level thatprecludes the mechanism listening to its own internal noise(reclimit_at). The mechanism produces output whenever the inputexceeds the adaptive threshold, and the output level is the differencebetween the input and the adaptive threshold. The adaptation andsuppression are coupled, and they jointly sharpen features like vowelformants which appear smeared in compressed BMM.

trise_at	Threshold rise rateDefault value: 10000. Upward Adaptation: This option specifies the rate at which theadaptive threshold will rise in response to a rise in signallevel. The default value, 10000, means that the adaptive thresholdresponds very quickly to increases in the input wave; essentially, itfollows the envelope of any rise in signal amplitude. Downward Adaptation: Following the cessation of sound, or a rapid dropin input level, temporal adaptation occurs in two stages as determinedby t1recovery_at, t2recovery_at and propt2t1_at: If the default valuesare used, the mechanism initially adapts at a rate slightly slowerthan the decay rate of the gammatone filter in the given channel, andthis represses much of the ringing of the impulse response of thefilter. Later the adaptation switches to a slower rate more in linewith data on auditory forward masking. The option propt2t1_atdetermines the point at which the initial fast rate of decay gives wayto the slower limiting decay rate.
t1recovery_at	The initial rate of threshold recovery relative to filter decay rateDefault value: 0.6. This option determines the initial rate of decay of the adaptivethreshold relative to the rate of decay of the auditory filter,provided propt2t1_at is less than unity. Values of t1recovery_at lessthan unity cause the adaptive threshold to decay more slowly than theauditory filter and thereby to remove the filter response from therepresentation when it is the sole reason for BMM activity. The rateof decay is linear with respect to the log-compressed BMM, so it islike an exponential decay with respect to the BMM.
t2recovery_at	The secondary threshold recovery rateDefault value: 0.2. This option determines the limiting rate of decay of the adaptivethreshold when the sound ceases provided propt2t1_at is less thanunity. The default value causes the adaptive threshold to decay moreslowly than the initial rate as observed in auditory forward masking.Note, however, that the system to this point is level independent,whereas auditory forward masking is level dependent.
propt2t1_at	The point at which t1recovery_at gives way to t2_recovery_atDefault value: 0.5. This option determines the point at which the initial fast rate ofdecay (t1recovery_at) gives way to the slower limiting decay rate(t2recovery_at). The nomanclature assumes that propt2t1_at is a valueless than unity. Otherwise the the roles of the initial and limitingdecays are reversed.
frecovery_at	Recovery rate across frequencyDefault value: 20. This parameter specifies the rate at which a threshold value in one channelpropagates to influence threshold in neighbouring channels.
reclimit_at	Limitation on recovery levelDefault units: mB. Default value: 500 mB. (mB=milliBells) In order to prevent the mechanism from encountering system noise,or alternately, to reduce sensitivity to stimulus noise, there is alimit placed on the recovery that the adaptive threshold can achieve.The limit, reclimit_at, is the limit of the sensitivity of the system.
gain_at	Output gainDefault units: scalar. Default value: 1.

The purpose of the Meddis module is to simulate neural transduction ofBMM as performed by the inner haircells of the cochlea (Meddis, 1986,1988). There is one haircell simulation unit for each output channelof the filterbank. The haircell equations (Meddis et al., 1990) aresolved using the wave digital filter algorithm described in Giguereand Woodland (1994). The characteristics of the haircell arecontrolled by options: fiber_med, thresh_med, and gain_med.

fiber_med	The spontaneous-rate of the simulated fiberDefault value: medium. Choices: medium, high. If medium is specified, a medium spontaneous-rate haircell fiber issimulated. If high is specified, a high spontaneous-ratefiber is simulated. The properties of these two types of fibersare listed in Table II in Meddis et al. (1990).The default value is medium.
thresh_med	The threshold shift of the fiberDefault Units: dB. Default value: 0. This option shifts the entire rate-intensity function of the haircellfiber horizontally to a higher or lower level, to accomodate changesin the scaling of the input wave. A positive (negative) valueincreases (decreases) the rate- and saturation-thresholds of the fiberby that amount. This operation does not change the dynamic range, thespontaneous and saturation rates, or the adaptation time constants orsynchronization index of the fiber.
gain_med	Output gainDefault units: scalar. Default value: 1. Note: There is an internal gain of 20.0 within the software ofthe Meddis haircell model itself. The total gain is therefore20.0 times the value for gain_med.

The following command generates the neural activity pattern using thegammatone auditory filterbank (the default) and adaptivethresholding (the default) for an input file named cegc:

> gennap cegc

The following command generates the neural activity pattern using thetransmission line filterbank and Meddis haircell transduction for cegc:

> gennap filter=tlf compress=off transduction=meddis cegc

The following command generates the neural activity pattern using thegammatone filterbank and Meddis haircelltransduction for input cegc:

> gennap compress=off transduction=meddis cegc

NOTE: docs/aimMeddisHewitt shows how to produce a Meddis and Hewitt(1991) model of pitch perception using the AIM software package, andhow to insert power compression between the gammatone filterbank andthe Meddis haircell bank.

REFERENCES

	Giguere, C. and Woodland, P.C. (1994). "A computational model ofthe auditory periphery for speech and hearing research," I. Ascendingpath. J.Acoust. Soc. Am. 95: 331-342.

FILES

.gennaprc	The options file for gennap.

genepn, gencgm, genbmm

Copyright (c) Applied Psychology Unit, Medical Research Council, 1995

Permission to use, copy, modify, and distribute this software without feeis hereby granted for research purposes, provided that this copyrightnotice appears in all copies and in all supporting documentation, and thatthe software is not redistributed for any fee (except for a nominalshipping charge). Anyone wanting to incorporate all or part of thissoftware in a commercial product must obtain a license from the MedicalResearch Council.

The MRC makes no representations about the suitability of thissoftware for any purpose. It is provided "as is" without express orimplied warranty.

THE MRC DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDINGALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT SHALLTHE A.P.U. BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGESOR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THISSOFTWARE.

The AIM software was developed for Unix workstations by JohnHoldsworth and Mike Allerhand of the MRC APU, under the direction ofRoy Patterson. The physiological version of AIM was developed byChristian Giguere. The options handler is by Paul Manson. The revisedSAI module is by Jay Datta. Michael Akeroyd extended the postscriptfacilites and developed the xreview routine for auditory imagecartoons.

The project was supported by the MRC and grants from the U.K. DefenseResearch Agency, Farnborough (Research Contract 2239); the EEC EspritBR Porgramme, Project ACTS (3207); and the U.K. Hearing Research Trust.