“ASR Status Report 2016-08-29”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
 
(5位用户的6个中间修订版本未显示)
第5行: 第5行:
 
|Zhiyuan Tang   
 
|Zhiyuan Tang   
 
||  
 
||  
*
+
*Test RNN residual learning along time, mostly trivial, hold on.
*
+
*LSTM 'C' component visualization[http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=tangzy&step=view_request&cvssid=557].
 
||  
 
||  
*  
+
*LSTM visualization, more components, more layers.
 
|-
 
|-
  
  
 
|-
 
|-
| YOUR NAME  
+
|Hang Luo  
 
||  
 
||  
* CONTENT
+
*Write book and finish the first version.
 
||  
 
||  
* CONTENT
+
*Modify the book.
 
|-
 
|-
  
  
  
 +
|-
 +
|Jingyi Lin 
 +
||
 +
* Write book
 +
# Finish the whole part of "direct graphic model" and half part of the "undirect ghraphic model".
 +
||
 +
* Continue with the book writing
 +
|-
 +
 +
|-
 +
|Ying Shi 
 +
||
 +
* nn visualization
 +
# some work on wsj (the length of the memory ,gender)[[http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=shiying&step=view_request&cvssid=556 here]]
 +
||
 +
* Continue research on lstm_c
 +
* daily report [[http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/Ying_Shi here]]
 +
|-
 +
 +
 +
|-
 +
|Lantian Li 
 +
||
 +
* Deep speaker embedding ( Random sample training RST ) [[http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=lilt&step=view_request&cvssid=564 here]].
 +
* Local training [[http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=shiying&step=view_request&cvssid=563 here]].
 +
||
 +
* Continue the two tasks.
 +
|-
 
|}
 
|}

2016年8月29日 (一) 04:34的最后版本

People Last Week This Week
Zhiyuan Tang
  • Test RNN residual learning along time, mostly trivial, hold on.
  • LSTM 'C' component visualization[1].
  • LSTM visualization, more components, more layers.
Hang Luo
  • Write book and finish the first version.
  • Modify the book.
Jingyi Lin
  • Write book
  1. Finish the whole part of "direct graphic model" and half part of the "undirect ghraphic model".
  • Continue with the book writing
Ying Shi
  • nn visualization
  1. some work on wsj (the length of the memory ,gender)[here]
  • Continue research on lstm_c
  • daily report [here]
Lantian Li
  • Deep speaker embedding ( Random sample training RST ) [here].
  • Local training [here].
  • Continue the two tasks.