1. Objective intelligibility measurements (3) (Lorenzo Picinali )


Subject: 1. Objective intelligibility measurements (3)
From:    Lorenzo Picinali  <LPicinali@xxxxxxxx>
Date:    Thu, 18 Sep 2008 11:05:42 +0100
List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>

Dear Kevin, I agree with what you say, and...yes, I think we have been using the adjective "objective" in a wrong way: talking about perception, it is true that "objective" shouldn't be used as we did. Nevertheless, I think it's important to distinguish an estimation done through statistical analysis of subjective tests' responses (the adjective "subjective" it is usually used for this kind of estimation), and an estimation done through analysis algorithms which should predict the response of the subjective tests: referring to this second case, the adjective "objective"it is usually used. In the case of speech intelligibility, we can have an estimation done measuring, for example, the SRT from a group of subject, and an estimation done through algorithms which should be able (even if they are not yet...) to predict the SRT that could be measured from the subjects: in the first case, I usually refer as a "subjective" estimation, whilst in the second case as an"objective" estimation. I understand this is not perfectly correct from a semantic point of view, but, from my (probably inadequate) knowledge of this field, this is how the two adjective are often used. Yours -- Lorenzo Picinali Music, Technology and Innovation Research Centre De Montfort University Clephan building, CL 0.19 The Gateway LE1 9BH Leicester UK tel. +44.0116.2551551, internal 6770 e-mail lpicinali@xxxxxxxx web http://www.mti.dmu.ac.uk/members-postgrad.htm


This message came from the mail archive
http://www.auditory.org/postings/2008/
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University