tag:blogger.com,1999:blog-8012401148587899501.comments2018-08-31T12:30:21.931+05:30Pronunciation EvaluationJames Salsmanhttp://www.blogger.com/profile/11470879525772180030noreply@blogger.comBlogger19125tag:blogger.com,1999:blog-8012401148587899501.post-23959396717521523122016-04-15T02:45:41.956+05:302016-04-15T02:45:41.956+05:30If you can't find the code at
http://pronunci...If you can't find the code at<br /> http://pronunciationeval.blogspot.com/2012/08/ronanki-gsoc-2012-pronunciation.html<br />or http://pronunciationeval.blogspot.com/2012/06/ronanki-gsoc-2012-pronunciation.html<br />then please try:<br />https://sourceforge.net/p/cmusphinx/code/HEAD/tree/branches/speecheval/ronanki/src/James Salsmanhttps://www.blogger.com/profile/11470879525772180030noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-22132706466533248672016-04-13T16:37:46.191+05:302016-04-13T16:37:46.191+05:30Hi Srikanth, the links do not work. :-(
Hi Srikanth, the links do not work. :-(<br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-3217596802290530322012-11-18T05:48:28.998+05:302012-11-18T05:48:28.998+05:30I am very sorry that I missed this blog from Troy ...I am very sorry that I missed this blog from Troy and his further updates on the CMU Sphinx blog: <a href="http://cmusphinx.sourceforge.net/2012/08/gsoc-2012-pronunciation-evaluation-troy-project-conclusions/" rel="nofollow">cmusphinx.sourceforge.net/2012/08/gsoc-2012-pronunciation-evaluation-troy-project-conclusions</a>James Salsmanhttps://www.blogger.com/profile/11470879525772180030noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-27942083653901444302012-08-27T23:53:42.884+05:302012-08-27T23:53:42.884+05:30Thank you very much! You guys did an amazing job =...Thank you very much! You guys did an amazing job =)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-24261510762092535292012-08-26T17:30:24.370+05:302012-08-26T17:30:24.370+05:30Sure, I didn't do proper documentation due to ...Sure, I didn't do proper documentation due to lack of time for coding till the last minute...I will write the documentation by the end of this week and will be provided at svn repository and will try to update at the wiki you mentioned as well.<br /><br />Thankssrikanth ronankihttps://www.blogger.com/profile/15384976912513321039noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-81563072748373514412012-08-22T03:20:07.496+05:302012-08-22T03:20:07.496+05:30Hi. I am really impressed by the work you have don...Hi. I am really impressed by the work you have done! I would like to be able to set up an educational website and use your technology to enable pronunciation scoring in it.<br /><br />I tried to do so by going through the files on the svn repository, but didn't succed. Do you have any plans to create some kind of documentation or to update the wiki at http://cmusphinx.sourceforge.net/wiki/faq#qhow_to_implement_pronunciation_evaluation ?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-31011332841516137902012-07-29T13:17:08.871+05:302012-07-29T13:17:08.871+05:30Date: 29/07/2012
Completed extracting mfcc, pncc,...Date: 29/07/2012<br /><br />Completed extracting mfcc, pncc, phonological featuressrikanth ronankihttps://www.blogger.com/profile/15384976912513321039noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-13666351972395337652012-07-26T00:42:00.108+05:302012-07-26T00:42:00.108+05:30Date: 25/07/2012
I completed the random phrase pr...Date: 25/07/2012<br /><br />I completed the random phrase pronunciation evaluation system and is in testing phase @ <a href="http://talknicer.net/~ronanki/test/" rel="nofollow">http://talknicer.net/~ronanki/test/</a>srikanth ronankihttps://www.blogger.com/profile/15384976912513321039noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-61568882152154325512012-07-24T01:37:59.690+05:302012-07-24T01:37:59.690+05:30This comment has been removed by the author.srikanth ronankihttps://www.blogger.com/profile/15384976912513321039noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-59011215567833179152012-07-24T01:37:20.598+05:302012-07-24T01:37:20.598+05:30Date: 23/07/2012
TIMIT statistics are ready !!
T...Date: 23/07/2012<br /><br />TIMIT statistics are ready !!<br /><br />Total 6300 utterances @ 10 utterances from each speaker (Total:630 speakers)<br />Successfully aligned files: 6149 out of 6300<br />and statistics are derived from those 6149 aligned files:<br /><br />http://talknicer.net/~ronanki/phrase_data/statistics/TIMIT_statistics.txt<br /><br />Position: 0/1/2 -> begin/middle/end<br />Count : Represents number of times each phone occurredsrikanth ronankihttps://www.blogger.com/profile/15384976912513321039noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-60998034513240061352012-07-23T01:15:22.988+05:302012-07-23T01:15:22.988+05:30Date: 22/07/2012
Still doing with random phrase p...Date: 22/07/2012<br /><br />Still doing with random phrase pronunciation evaluation @ http://talknicer.net/~ronanki/test/random.html and testing PNCC on TIMIT database !!srikanth ronankihttps://www.blogger.com/profile/15384976912513321039noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-3908482085601156242012-07-20T00:51:00.336+05:302012-07-20T00:51:00.336+05:30Date: 19/07/2012
Studied about Power Normalized C...Date: 19/07/2012<br /><br />Studied about Power Normalized Cepstral Coefficients (PNCC) which are more robust towards speech recognition even in noisy environment. PNCC are 13 in dimension, computationally more cost than MFCC but performs better than MFCC in speech recognition. Got the code, able to run it. Will Compare accuracy on TIMIT database tomorrow !!srikanth ronankihttps://www.blogger.com/profile/15384976912513321039noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-59342231881659923582012-07-19T23:07:03.716+05:302012-07-19T23:07:03.716+05:30I somehow forgot this...let me start from today on...I somehow forgot this...let me start from today onwards...!!srikanth ronankihttps://www.blogger.com/profile/15384976912513321039noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-24289162032453902832012-05-31T03:08:38.760+05:302012-05-31T03:08:38.760+05:30The decision to move from C to Java for Sphinx 4 w...The decision to move from C to Java for Sphinx 4 was made a decade ago and the rationale is detailed in http://cmusphinx.sourceforge.net/wiki/sphinx4:february2122meeting Sphinx3 is still maintained, according to http://cmusphinx.sourceforge.net/wiki/versions but it is very stable. Thank you for your kind words.James Salsmanhttps://www.blogger.com/profile/11470879525772180030noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-87873255852233776592012-05-31T02:51:26.249+05:302012-05-31T02:51:26.249+05:30Thanks for the explanation. I didn't know that...Thanks for the explanation. I didn't know that sphinx3 had been implemented in C. Speaking of which, why the decision to completely change the language from one version to another? And is there any development in version 3 still or has it completely lost support?<br /><br />I am very happy to know that you are thinking of creating stand-alone versions for mobile devices. I think the work you are developing is very important! Congratulations =)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-52302825438481608982012-05-31T01:27:24.776+05:302012-05-31T01:27:24.776+05:30I can answer that, as it was my decision as the pr...I can answer that, as it was my decision as the project mentor. Sphinx3 is written in C, while Sphinx4 is in Java, which takes much more memory and time to run. We hope to create stand-alone versions which run in limited memory on mobile devices and OLPC laptops. The Viterbi beam search algorithm is very memory bus bandwidth intensive, so using the Java Virtual Machine instead of native code, and the memory overhead which Java data structures entail (not to mention managed memory such as unpredictable garbage collection) would be completely impractical for response time, memory utilization, and battery life on mobile and OLPC systems.<br /><br />Furthermore, Sphinx4 currently has some problems doing forced alignment of phonemes:<br />http://sourceforge.net/projects/cmusphinx/forums/forum/382337/topic/4720560<br />http://sourceforge.net/projects/cmusphinx/forums/forum/382337/topic/4740193James Salsmanhttps://www.blogger.com/profile/11470879525772180030noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-15725478177476715652012-05-31T00:31:44.192+05:302012-05-31T00:31:44.192+05:30Sorry for my ignorance, but why do you use shpinx3...Sorry for my ignorance, but why do you use shpinx3 and not the latest version (4)?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-31276611227970026002012-05-19T03:12:44.719+05:302012-05-19T03:12:44.719+05:30Approved! By the OLPC Contributors Program! \o/Approved! By the OLPC Contributors Program! \o/James Salsmanhttps://www.blogger.com/profile/11470879525772180030noreply@blogger.comtag:blogger.com,1999:blog-8012401148587899501.post-49881847916584906022012-04-25T17:34:59.121+05:302012-04-25T17:34:59.121+05:30Looks like a nice project for our XO from laptop.o...Looks like a nice project for our XO from laptop.org<br />If you are interested in get hardware to work on this, just ask.Anonymoushttps://www.blogger.com/profile/02412149415810582119noreply@blogger.com