<div dir="ltr">Wouldn't that be even harder to do? Spectrograms are not very clear, and reading them from scratch (without the sound) is a very difficult task, even for a human. </div><div class="gmail_extra"><br><br><div class="gmail_quote">
2013/3/6 WHITELOCK, Pete <span dir="ltr"><<a href="mailto:pete.whitelock@oup.com" target="_blank">pete.whitelock@oup.com</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div lang="EN-GB" link="blue" vlink="purple"><div><p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d">Isn’t everyone missing the point here. It’s not about ASR; Mario asked about automatic analysis of spectrograms, a problem in visual pattern extraction.<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p><p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:#1f497d">Pete Whitelock, PhD</span><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><br>
</span><span style="font-size:10.0pt;font-family:"Courier New";color:#1f497d">Principal Language Engineer, Technology<u></u><u></u></span></p><p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New";color:#1f497d">Academic Dictionaries</span><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"> <br>
</span><span style="font-size:10.0pt;font-family:"Courier New";color:#1f497d">Oxford University Press</span><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1f497d"><u></u> <u></u></span></p><p class="MsoNormal"><b><span lang="EN-US" style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">From:</span></b><span lang="EN-US" style="font-size:10.0pt;font-family:"Tahoma","sans-serif""> <a href="mailto:corpora-bounces@uib.no" target="_blank">corpora-bounces@uib.no</a> [mailto:<a href="mailto:corpora-bounces@uib.no" target="_blank">corpora-bounces@uib.no</a>] <b>On Behalf Of </b>Jim Fidelholtz<br>
<b>Sent:</b> 06 March 2013 15:18<br><b>To:</b> Matías Guzmán<br><b>Cc:</b> Mario Crespo Miguel; <a href="mailto:corpora-request@uib.no" target="_blank">corpora-request@uib.no</a>; <a href="mailto:corpora@uib.no" target="_blank">corpora@uib.no</a><br>
<b>Subject:</b> Re: [Corpora-List] International Phonetic Alphabet transcription tool / software<u></u><u></u></span></p><div><div class="h5"><p class="MsoNormal"><u></u> <u></u></p><div><p class="MsoNormal">Hi, Mario,<u></u><u></u></p>
<div><p class="MsoNormal"><u></u> <u></u></p></div><div><p class="MsoNormal">My advice is to take Matías' comments very seriously. Over 40 years ago, I was involved with the US government group which funded most American research in speech recognition. Many very bright researchers were involved in this research. Unfortunately, despite the fact that even then (about 1970 and just earlier) there were 'effective' systems trainable on small vocabularies (eg, numbers 0-9 for the Post Office) and individual speakers, the field was populated by many charlatans/snake-oil salesman types, and believe me, I use those terms advisedly. Currently, although I am very aware of the multiple orders of magnitude increases in computer speed, capacity and general gee-whiz factor, I am still not convinced of the general applicability of such systems, and of course even less convinced of any linguistic interest in such systems, even for general phone answering, etc., and of course (unfortunately) much less for applications like what you are asking about. <u></u><u></u></p>
</div><div><p class="MsoNormal"><u></u> <u></u></p></div><div><p class="MsoNormal">The usual disclaimers apply, and you are unlikely to encounter anyone quite as negative as I am about speech recognition, and I'm sure the field still contains many brilliant, ethical researchers, and I have not been following publications in the field in recent decades. Nevertheless, I stand by my opinions, based on well-informed experience, even though it was decades ago. Example: my current experience with automatic speech-recognition systems is that, more often than not, I end up being transferred to a human operator, and I speak Midwestern American English, which should be relatively easy for these systems to recognize, from a linguistically-informed standpoint. Sorry to be so negative, but I'd bet lots of money that you won't be successful in finding anything which would be very satisfying, even if you have program-tweaking powers. I would say that if you have lowered expectations and an extremely liberal attitude, you *might* find something which could give you some minimum help in your research. Perhaps it would be better to be (or use/hire) an experienced dialectologist, design very intelligent experiments which could get useful results from minimal recordings, and do your own transcriptions. I know that's a tall order. Good luck.<u></u><u></u></p>
</div><div><p class="MsoNormal"><u></u> <u></u></p></div><div><p class="MsoNormal">Jim<u></u><u></u></p></div></div><div><p class="MsoNormal" style="margin-bottom:12.0pt"><u></u> <u></u></p><div><p class="MsoNormal">On Wed, Mar 6, 2013 at 8:45 AM, Matías Guzmán <<a href="mailto:mortem.dei@gmail.com" target="_blank">mortem.dei@gmail.com</a>> wrote:<u></u><u></u></p>
<div><p class="MsoNormal">Mario, I doubt there is anything that can do what you want. As I understand it, speech recognition systems depend to a great degree of a language model that predicts what word could come next, and then try to match what the speaker said to a database. I don't see how a program could transcribe for you if the /t/ is dental or alveolar, I don't even think speech recognition software do a decent job working on single syllables. But I'm not an expert, maybe someone can correct me.<u></u><u></u></p>
<div><p class="MsoNormal"><u></u> <u></u></p></div><div><p class="MsoNormal">Regards, Matías<u></u><u></u></p></div></div><div><p class="MsoNormal" style="margin-bottom:12.0pt"><u></u> <u></u></p><div><div><div><p class="MsoNormal">
2013/3/6 Mario Crespo Miguel <<a href="mailto:mario.crespo@uca.es" target="_blank">mario.crespo@uca.es</a>><u></u><u></u></p></div></div><blockquote style="border:none;border-left:solid #cccccc 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<div><div><div><p style="margin:0cm;margin-bottom:.0001pt"><span lang="EN-US" style="font-size:10.5pt;font-family:"Verdana","sans-serif"">Dear members of corporalist, </span><u></u><u></u></p><p style="margin:0cm;margin-bottom:.0001pt">
<span lang="EN-US" style="font-size:10.5pt;font-family:"Verdana","sans-serif"">In the research of phonetic dialectology it is extremely important to be able to differentiate between the different sounds pronounced by the subjects being studied and be able to transcribe them according to linguistic criteria. This task can be extremely exhausting and subjective for the researcher when carried out without any help. </span><u></u><u></u></p>
<p style="margin:0cm;margin-bottom:.0001pt"><span lang="EN-US" style="font-size:10.5pt;font-family:"Verdana","sans-serif"">I wonder if you know a tool or software able to represent and analyze spectrograms of sounds pronounced and transcribe such sounds into the International Phonetic Alphabet (IPA). </span><u></u><u></u></p>
<p style="margin:0cm;margin-bottom:.0001pt"><span lang="EN-US" style="font-size:10.5pt;font-family:"Verdana","sans-serif"">Thank you very much in advance, </span><u></u><u></u></p><p style="margin:0cm;margin-bottom:.0001pt">
<span lang="EN-US" style="font-size:10.5pt;font-family:"Verdana","sans-serif"">Mario Crespo</span><u></u><u></u></p></div><p class="MsoNormal"><u></u> <u></u></p></div></div><p class="MsoNormal" style="margin-bottom:12.0pt">
_______________________________________________<br>UNSUBSCRIBE from this page: <a href="http://mailman.uib.no/options/corpora" target="_blank">http://mailman.uib.no/options/corpora</a><br>Corpora mailing list<br><a href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a><br>
<a href="http://mailman.uib.no/listinfo/corpora" target="_blank">http://mailman.uib.no/listinfo/corpora</a><u></u><u></u></p></blockquote></div><p class="MsoNormal"><u></u> <u></u></p></div><p class="MsoNormal" style="margin-bottom:12.0pt">
<br>_______________________________________________<br>UNSUBSCRIBE from this page: <a href="http://mailman.uib.no/options/corpora" target="_blank">http://mailman.uib.no/options/corpora</a><br>Corpora mailing list<br><a href="mailto:Corpora@uib.no" target="_blank">Corpora@uib.no</a><br>
<a href="http://mailman.uib.no/listinfo/corpora" target="_blank">http://mailman.uib.no/listinfo/corpora</a><u></u><u></u></p></div><p class="MsoNormal"><br><br clear="all"><u></u><u></u></p><div><p class="MsoNormal"><u></u> <u></u></p>
</div><p class="MsoNormal">-- <br>James L. Fidelholtz<br>Posgrado en Ciencias del Lenguaje<br>Instituto de Ciencias Sociales y Humanidades<br>Benemérita Universidad Autónoma de Puebla, MÉXICO <u></u><u></u></p></div></div>
</div></div>
<p>Oxford University Press (UK) Disclaimer</p>
<p>This message is confidential. You should not copy it or disclose its contents
to anyone. You may use and apply the information for the intended purpose only.
OUP does not accept legal responsibility for the contents of this message. Any
views or opinions presented are those of the author only and not of OUP. If this
email has come to you in error, please delete it, along with any attachments.
Please note that OUP may intercept incoming and outgoing email
communications.</p>
</div>
</blockquote></div><br></div>