“ASR Status Report 2016-11-21”版本间的差异

• fighting with kazak speech recognition system:because the huge size of HCLG.fst the decoding job always make the sever done.

There are several method I have tried

• change the size or word list and corpus this method not worked very well
• prune the LM .And the parameter used to prune the LM is 2e-7 the size of LM reduce from 290M to 60M but the result about wer is very poor
• I have upload some result about several experiment to CVSS (I ought to put a link here but I modify the page is my dormitory so I can't visit CVSS so...)

• there are too much private affairs so the job about visualization last week has been delayed I will try my best to finish it the week

• make a report of 2 Inter Speech papers
• run joint training experiments and the result is [1]

• read papers about highway connection and multi-task.

• paper reading kazak speech recognition data prep

• baseline of kazak speech recoginition

• almost finished the additinal experiments of joint learning (speech & spk) for taslp (multi-task, ivector as part of input)[2].

• report for Weekly Reading;
• joint training for bilingual.