“ASR Status Report 2017-1-3”版本间的差异

来自cslt Wiki
跳转至： 导航、 搜索
 
（3位用户的4个中间修订版本未显示）
第1行： 第1行：
 +
  
  
第4行： 第5行：
 
!Date!!People !! Last Week !! This Week
 
!Date!!People !! Last Week !! This Week
 
|-
 
|-
−
| rowspan="7"|2016.12.26
+
| rowspan="7"|2017.1.3
  
  
第22行： 第23行：
 
||  
 
||  
 
* finish demo of distraction task
 
* finish demo of distraction task
−
** choose subject / SVM type & train model in data sender:[[dataSender.png]| data sender]
+
** choose subject / SVM type & train model in data sender: [[媒体文件:dataSender.png|data sender]]
−
** check the status of driver in data analyzer:
+
** check the status of driver in data analyzer: [[媒体文件:dataAna1.png|focused]] [[媒体文件:dataAna2.png|distracted]]
−
* write a document on the mechanism of the program of both the data sender and data analyzer
+
* write a document on the mechanism of the program of both the data sender and data analyzer '''[delivery]'''
 
||  
 
||  
 
*  
 
*  
第35行： 第36行：
 
|Hang Luo
 
|Hang Luo
 
||  
 
||  
−
* TRP for joint training
+
* TRP for joint training '''[delivery]'''
 
* Make a review of mix-lingual
 
* Make a review of mix-lingual
 
||  
 
||  
−
* Make a conclude for aurora4 and thchs30 joint training
+
*  
 
|-
 
|-
  
第45行： 第46行：
 
|Ying Shi   
 
|Ying Shi   
 
||  
 
||  
−
* TRP for kazak speech recognition
+
* TRP for kazak speech recognition '''[delivery]'''
 
* make new test set which have lower ppl on LM
 
* make new test set which have lower ppl on LM
 
* The characteristics of the Kazak Language summarized by nurpolat [http://cslt.riit.tsinghua.edu.cn/mediawiki/images/8/88/%E5%93%88%E8%90%A8%E5%85%8B%E8%AF%AD%E8%A8%80%E7%89%B9%E7%82%B9.pdf here]
 
* The characteristics of the Kazak Language summarized by nurpolat [http://cslt.riit.tsinghua.edu.cn/mediawiki/images/8/88/%E5%93%88%E8%90%A8%E5%85%8B%E8%AF%AD%E8%A8%80%E7%89%B9%E7%82%B9.pdf here]
第81行： 第82行：
 
|Zhiyuan Tang  
 
|Zhiyuan Tang  
 
||  
 
||  
−
* TRP of "How to Config Kaldi nnet3 (in Chinese)", describe the usage of 34 components;
+
* TRP of "How to Config Kaldi nnet3 (in Chinese)", describe the usage of 34 components '''[delivery]''';
 
* additional notes on "Multi-task Recurrent Model for True Multilingual Speech Recognition";
 
* additional notes on "Multi-task Recurrent Model for True Multilingual Speech Recognition";
 
* Generative models, part of Chapter Deep Learning.
 
* Generative models, part of Chapter Deep Learning.

2017年1月4日 (三) 07:26的最后版本


Date People Last Week This Week
2017.1.3


Jingyi Lin
  • Modify and supply Dr.Wang's personal page.
  • Prepare for the Annual Meeting.
  • Check CSLT books.(2,5)
  • Check CSLT books.
Yanqing Wang
  • finish demo of distraction task
  • write a document on the mechanism of the program of both the data sender and data analyzer [delivery]
Hang Luo
  • TRP for joint training [delivery]
  • Make a review of mix-lingual
Ying Shi
  • TRP for kazak speech recognition [delivery]
  • make new test set which have lower ppl on LM
  • The characteristics of the Kazak Language summarized by nurpolat here
  • new kazak AM and LM
Yixiang Chen
  • Prepare the thesis proposal
  • Integrate CNN + max-margin.
  • Integrate CNN + max-margin.
Lantian Li
  • Deep speaker embedding
  • Write book of robustness SRE.
  • Wechat open account -- Hole.
  • Replay detection on INTERSPEECH chanllenge -- Get the database.
  • Deep speaker embedding.
  • Write book.
  • Replay detection on INTERSPEECH chanllenge.
Zhiyuan Tang
  • TRP of "How to Config Kaldi nnet3 (in Chinese)", describe the usage of 34 components [delivery];
  • additional notes on "Multi-task Recurrent Model for True Multilingual Speech Recognition";
  • Generative models, part of Chapter Deep Learning.
  • check and submit the above 3 writings.




Date People Last Week This Week
2016.12.26


Jingyi Lin
  • Learn and make Dr.Wang's personal web page.
  • Prepare for the CSLT's Annual Meeting.
  • Finish Dr.Wang's personal web page.
  • Take photos for menmbers in CSLT.
Yanqing Wang
  • implement the detection mechanism by socket
  • find best parameters to avoid over-fitting
  • add two-class-SVM to the program
  • make GUI more pretty and easy to use
  • improve the program's robustness
  • screenshot:
  • write a document on the program
Hang Luo
  • Run joint training and write systemic script and documents
  • Finish joint training documents
  • Conclude joint training experiments result
  • Make a review on mixlingual
Ying Shi
  • crawl corpus from internet.(I don't know whether the corpus is right or not.......)
  • make new LM(complete)
  • train new AM(complete)
  • a part of TRP
  • finish the TRP
Yixiang Chen
  • Prepare the input of speech data (trick of block segmentation)
  • Complete the init version on max-margin SRE.
  • Write TRP-20160012 "基于Kaldi i-vector的说话人识别系统使用说明".
  • Prepare the thesis proposal.
  • Integrate CNN + max-margin.
Lantian Li
  • Deep speaker embedding
    • Prepare two datasets and make the i-vector baselines.
  • Write TRP-20160012 "基于Kaldi i-vector的说话人识别系统使用说明".
  • Write book of robustness SRE.
  • Wechat open account.
  • Deep speaker embedding.
  • Write book.
  • Replay detection on INTERSPEECH chanllenge.
Zhiyuan Tang
  • TRP of "How to Config Kaldi nnet3 (in Chinese)", not finished yet;
  • outline of TRP for "Multi-task Recurrent Model for True Multilingual Speech Recognition";
  • Generative models, part of Chapter Deep Learning.
  • Finish the above 3 writings.
取自“http://index.cslt.org/mediawiki/index.php?title=ASR_Status_Report_2017-1-3&oldid=25378”