<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style></head>
<body class='hmmessage'><div dir='ltr'>
<font size="3"><font style="font-size: 10pt;" size="2">Dear pranavshriram</font></font><BR><font size="3"><font size="2"></font></font> <BR><font size="3"><font size="2">first thank you for replied. Actually, I have read the two papers( espcially the<font size="3"> </font>second one), but I will read them again to figure out the other things to include.</font></font><BR><font size="3"></font> <BR><font size="3"><font size="2">for the phone decoder you mean that the output of the recognizer(decoder) will be in the phone level( not a word level).</font></font><BR><font size="3" face="Times New Roman"><font size="2" face="Tahoma">thank you.</font></font><BR><font size="3" face="Times New Roman"><font size="2" face="Tahoma"></font></font> <BR><font size="3" face="Times New Roman"><font size="2" face="Tahoma">Bassam.</font><BR></font><br><br> <BR><p align="center"><font color="#494429" size="3" face="Times New Roman"></font> </p>
<BR><div><hr id="stopSpelling">From: pranavshriram@gmail.com<br>Date: Tue, 10 Jul 2012 09:29:34 +0530<br>Subject: Re: [SRILM User List] Confidence Measure<br>To: bassam_qatab@hotmail.com<br>CC: srilm-user@speech.sri.com<br><br><div class="ecxgmail_quote"><blockquote style="padding-left: 1ex; border-left-color: rgb(204, 204, 204); border-left-width: 1px; border-left-style: solid;" class="ecxgmail_quote"><div><div dir="ltr"><font size="3" face="Times New Roman"><span style='line-height: 115%; font-family: "Times New Roman","serif"; font-size: 14pt;'>Or should
I have to calculate the confidence measure?</span></font></div></div></blockquote><div>Word posterior probability computed using confusion network itself IS a confidence measure. But it is not the only one, there are many others too. </div>
<div>e.g. see </div><div>[] H Jiang, “Confidence measures for speech recognition: A survey”, Speech Communication, 2005</div><div>[] F. Wessel, R. Schlüter, K. Macherey and H. Ney, “Confidence measures for large vocabulary continuous speech recognition”, IEEE Transactions on Speech and Audio Processing , 2001</div>
<div><br></div><div>For pronunciation evaluation at phone-level, you may need to compute phone posterior probability using a phone decoder.</div><div><br></div></div>-- <br><div dir="ltr">*<br>*<br><font face="georgia, bookman old style, palatino linotype, book antiqua, palatino, trebuchet ms, helvetica, garamond, sans-serif, arial, verdana, avante garde, century gothic, comic sans ms, times, times new roman, serif">The best way to get something done is to begin. ~Author Unknown</font><br>
<br></div><br></div> </div></body>
</html>