Subject: Auditory Illusions (Imran Dhamani )


Subject: Subject: Auditory Illusions
From:    Imran Dhamani  <imrandhamani@xxxxxxxx>
Date:    Wed, 3 Aug 2011 10:34:40 +0530
List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>

--0-633072965-1312347880=:34678 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Dear Nedra,=0AI have a slightly different=0Aunderstanding in the regard of = auditory illusions such as the Mc Gurk effect=0Anot contributing to hearing= aid design. I think these illusions may have more=0Ato contribute to not o= nly hearing aid design but also to psycho-acoustics. One of the key uses of= the Mc=0AGurk effect was in terms of studying the re-organisation/remappin= g process in=0Athe auditory and visual cortex which has been recently repor= ted due to auditory=0Adeprivation (deafness) in hearing impaired children a= nd the changes to the same=0Awith cochlear implantation and I guess the sam= e would apply to hearing aids as=0Awell (especially the ones with severe he= aring loss). One of the other=0Ainteresting illusions (although not purely = an auditory illusion =E2=80=93 or maybe with=0Athe current discussion that = I have noticed in the list may be not even an=0Aillusion) I remember workin= g with hearing aid and cochlear implanted users is=0Athe Kiki-Bouba effect = which helps us study the sound-shape correspondences in=0Athe brain. One of= the key missing link I think in the current wave of hearing=0Aaid technolo= gy which is mainly driven via psycho-acoustical difficulties that a=0Aheari= ng impaired user faces is the overall auditory processing (central auditory= =0Aprocessing as well as Auditory- visual processing) difficulties that a h= earing=0Aloss impinges along-with the lower level psycho-acoustic deficits = like poor=0Afrequency selectivity, temporal resolution etc. Maybe broadenin= g the definition=0Aof psycho-acoustics a bit these aspects may also be well= under its domain. The=0Asubtle but existing difference between two hearing= aid users with the same type=0Aof hearing loss characteristics may well be= in terms of their central auditory=0Avisual processing abilities which mig= ht have been differently affected by the=0Aloss (although there may be alte= rnative aspects such as difference in some=0Arelatively less explored low l= evel psycho-acoustic abilities adding to it). In a=0Anutshell what I want t= o express is that studying these illusions is to somehow=0Aensure that we r= each limitations which are imposed by an impaired ear and not=0Athose impos= ed by inadequate amplification systems and may be try to find out=0Aand com= pensate for the additional auditory-visual processing/integration limitatio= ns that=0Athe hearing loss has added using appropriate rehabilitative/train= ing measures. =0A=0AI don=E2=80=99t know why there were those=0Astrange cha= racters visible in my previous post for this topic so I thought of=0Aattach= ing that previous message at the end as a paragraph for better=0Areadabilit= y. Apologies to all for any inconvenience caused. =0A=C2=A0=0AAlthough I am= not an expert in=0Athis area, but I have done some preliminary (unpublishe= d) work on the use of=0Athe Mc Gurk effect for the evaluation of current da= y multichannel digital=0Ahearing aids. Few of the basic assumptions for the= rationale and implications=0Awere as follows. One of the main benefits of = audio visual integration is in=0Adifficult listening situations when either= the speech is degraded or there is=0Abackground noise. The visual cue in a= n auditory visual integration task is=0Arelatively unaffected by noise. Hea= ring impaired listeners especially those=0Awith relatively larger auditory = deprivation periods rely more on visual cues=0A(the AV balance is slightly = tilted towards visual dominance) in speech=0Aperception than normal hearing= listeners, thus making them better speech=0Areaders and also relatively po= or AV integrators. The psycho-acoustical aspects=0Aof sensory-neural hearin= g loss suggest a reduction in the possibility of=0Aperceiving certain class= es of speech sounds especially in presence of noise. Moreover=0Asome of the= speech sounds may not be heard in noisy environments no matter how=0Ainten= sely they are amplified by the hearing aids. The Digital multichannel=0Ahea= ring aids may also have some amount of internal distortion and delay due to= =0Adigital processing and filtering. In such scenarios the hearing aid user= may=0Abenefit from the visual cues provided by a speakers facial and lip m= ovements which=0Ashould relatively be unaffected by noise and thus boost sp= eech perception for=0Aenergetic masking at poor signal to noise ratios as w= ell as in informational=0Amasking. We presented hearing aid users (relative= ly homogenous group in terms=0Aof hearing aid used) the Mc Gurk stimuli in = both congruent and in-congruent=0Aconditions in quiet and in noise (three d= ifferent SNRs) at comfortable level at=0Awhich the subjects scored > 70% on= a screening test using PB words. A=0Acriterion of 3/4th fusion responses w= as kept to determine the presence of Mc=0AGurk effect. The results indicate= d that subjects with normal hearing performed=0Abetter than the subjects us= ing HA=E2=80=99s in all conditions. The presence of auditory=0Aand visual i= nformation simultaneously in the congruent condition was beneficial=0Afor s= peech perception in quiet and in noise. Noise reduced the subjects ability= =0Ato perceive speech at poor SNR and had a more severe impact on the perfo= rmance=0Aof the HA users than in normal subjects. The Mc Gurk effect was ab= sent in the=0AHA users at the poorer SNRs compared to normal listeners and = in these=0Aconditions the HA users responses to the Mc Gurk stimuli were ma= inly visually=0Adominated. Background noise and increased listening effort = are significant=0Afactors influencing hearing-aid satisfaction and one of t= he major reason for=0Arejection of HA=E2=80=99s. Testing Mc Gurk effect in = a noisy environment may be a=0Auseful way to understand auditory visual spe= ech perception in HA users and=0Averify the benefits of aided AV speech per= ception in noise in HA users. Some of=0Athe implications in terms of hearin= g aid rehabilitation for the same may be in=0Aterms of enhancing optimal HA= fitting to achieve not only good auditory=0Aperception in noise but also o= ptimum auditory visual perception in noise and=0Athe emphasis on auditory t= raining and the use of speech reading skills. The=0Astudy that I have menti= oned above was in no way devoid of limitations like=0Alesser sample size an= d the hearing aids were used at the same programs that the=0AHA users were = using in their everyday listening environments and thus all the=0Auser had = different program settings. May be the rationale and implications=0Amight b= e of some interest to you though.=0A=C2=A0=0ARegards,=0AImran Dhamani --0-633072965-1312347880=:34678 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable <html><body><div style=3D"color:#000; background-color:#fff; font-family:ti= mes new roman, new york, times, serif;font-size:12pt"><div class=3D"yiv5809= 31635MsoNormal" style=3D"text-align:justify;">Dear Nedra,</div>=0A=0A<div c= lass=3D"yiv580931635MsoNormal" style=3D"text-align:justify;"><span style=3D= ""></span>I have a slightly different=0Aunderstanding in the regard of audi= tory illusions such as the Mc Gurk effect=0Anot contributing to hearing aid= design. I think these illusions may have more=0Ato contribute to not only = hearing aid design but also to psycho-acoustics. One of the key uses of the= Mc=0AGurk effect was in terms of studying the re-organisation/remapping pr= ocess in=0Athe auditory and visual cortex which has been recently reported = due to auditory=0Adeprivation (deafness) in hearing impaired children and t= he changes to the same=0Awith cochlear implantation and I guess the same wo= uld apply to hearing aids as=0Awell (especially the ones with severe hearin= g loss). One of the other=0Ainteresting illusions (although not purely an a= uditory illusion =E2=80=93 or maybe with=0Athe current discussion that I ha= ve noticed in the list may be not even an=0Aillusion) I remember working wi= th hearing aid and cochlear implanted users is=0Athe Kiki-Bouba effect whic= h helps us study the sound-shape correspondences in=0Athe brain. One of the= key missing link I think in the current wave of hearing=0Aaid technology w= hich is mainly driven via psycho-acoustical difficulties that a=0Ahearing i= mpaired user faces is the overall auditory processing (central auditory=0Ap= rocessing as well as Auditory- visual processing) difficulties that a heari= ng=0Aloss impinges along-with the lower level psycho-acoustic deficits like= poor=0Afrequency selectivity, temporal resolution etc. Maybe broadening th= e definition=0Aof psycho-acoustics a bit these aspects may also be well und= er its domain. The=0Asubtle but existing difference between two hearing aid= users with the same type=0Aof hearing loss characteristics may well be in = terms of their central auditory=0Avisual processing abilities which might h= ave been differently affected by the=0Aloss (although there may be alternat= ive aspects such as difference in some=0Arelatively less explored low level= psycho-acoustic abilities adding to it). In a=0Anutshell what I want to ex= press is that studying these illusions is to somehow=0Aensure that we reach= limitations which are imposed by an impaired ear and not=0Athose imposed b= y inadequate amplification systems and may be try to find out=0Aand compens= ate for the additional auditory-visual processing/integration limitations t= hat=0Athe hearing loss has added using appropriate rehabilitative/training = measures. =0A=0A</div><div class=3D"yiv580931635MsoNormal" style=3D"text-al= ign:justify;"><br>I don=E2=80=99t know why there were those=0Astrange chara= cters visible in my previous post for this topic so I thought of=0Aattachin= g that previous message at the end as a paragraph for better=0Areadability.= Apologies to all for any inconvenience caused. </div>=0A=0A<div class=3D"y= iv580931635MsoNormal" style=3D"text-align:justify;">&nbsp;</div>=0A=0A<div = class=3D"yiv580931635MsoNormal" style=3D"text-align:justify;">Although I am= not an expert in=0Athis area, but I have done some preliminary (unpublishe= d) work on the use of=0Athe Mc Gurk effect for the evaluation of current da= y multichannel digital=0Ahearing aids. Few of the basic assumptions for the= rationale and implications=0Awere as follows. One of the main benefits of = audio visual integration is in=0Adifficult listening situations when either= the speech is degraded or there is=0Abackground noise. The visual cue in a= n auditory visual integration task is=0Arelatively unaffected by noise. Hea= ring impaired listeners especially those=0Awith relatively larger auditory = deprivation periods rely more on visual cues=0A(the AV balance is slightly = tilted towards visual dominance) in speech=0Aperception than normal hearing= listeners, thus making them better speech=0Areaders and also relatively po= or AV integrators. The psycho-acoustical aspects=0Aof sensory-neural hearin= g loss suggest a reduction in the possibility of=0Aperceiving certain class= es of speech sounds especially in presence of noise. Moreover=0Asome of the= speech sounds may not be heard in noisy environments no matter how=0Ainten= sely they are amplified by the hearing aids. The Digital multichannel=0Ahea= ring aids may also have some amount of internal distortion and delay due to= =0Adigital processing and filtering. In such scenarios the hearing aid user= may=0Abenefit from the visual cues provided by a speakers facial and lip m= ovements which=0Ashould relatively be unaffected by noise and thus boost sp= eech perception for=0Aenergetic masking at poor signal to noise ratios as w= ell as in informational=0Amasking. We presented hearing aid users (relative= ly homogenous group in terms=0Aof hearing aid used) the Mc Gurk stimuli in = both congruent and in-congruent=0Aconditions in quiet and in noise (three d= ifferent SNRs) at comfortable level at=0Awhich the subjects scored &gt; 70%= on a screening test using PB words. A=0Acriterion of 3/4th fusion response= s was kept to determine the presence of Mc=0AGurk effect. The results indic= ated that subjects with normal hearing performed=0Abetter than the subjects= using HA=E2=80=99s in all conditions. The presence of auditory=0Aand visua= l information simultaneously in the congruent condition was beneficial=0Afo= r speech perception in quiet and in noise. Noise reduced the subjects abili= ty=0Ato perceive speech at poor SNR and had a more severe impact on the per= formance=0Aof the HA users than in normal subjects. The Mc Gurk effect was = absent in the=0AHA users at the poorer SNRs compared to normal listeners an= d in these=0Aconditions the HA users responses to the Mc Gurk stimuli were = mainly visually=0Adominated. Background noise and increased listening effor= t are significant=0Afactors influencing hearing-aid satisfaction and one of= the major reason for=0Arejection of HA=E2=80=99s. Testing Mc Gurk effect i= n a noisy environment may be a=0Auseful way to understand auditory visual s= peech perception in HA users and=0Averify the benefits of aided AV speech p= erception in noise in HA users. Some of=0Athe implications in terms of hear= ing aid rehabilitation for the same may be in=0Aterms of enhancing optimal = HA fitting to achieve not only good auditory=0Aperception in noise but also= optimum auditory visual perception in noise and=0Athe emphasis on auditory= training and the use of speech reading skills. The=0Astudy that I have men= tioned above was in no way devoid of limitations like=0Alesser sample size = and the hearing aids were used at the same programs that the=0AHA users wer= e using in their everyday listening environments and thus all the=0Auser ha= d different program settings. May be the rationale and implications=0Amight= be of some interest to you though.</div>=0A=0A<div class=3D"yiv580931635Ms= oNormal" style=3D"text-align:justify;">&nbsp;</div>=0A=0A<div class=3D"yiv5= 80931635MsoNormal" style=3D"text-align:justify;">Regards,</div>=0A=0A<div c= lass=3D"yiv580931635MsoNormal" style=3D"text-align:justify;">Imran Dhamani<= /div><div>&nbsp;</div><div><br></div></div></body></html> --0-633072965-1312347880=:34678--


This message came from the mail archive
/var/www/postings/2011/
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University