“Asr-progress 2016.10”版本间的差异

来自cslt Wiki
跳转至： 导航、 搜索
（以“ ===Daily Report=== {|class="wikitable" ! Date !! Person !! start!! leave !! hours ||status |- | rowspan="3"|2016.10.17 |Ying Shi || || || || |- |Hang...”为内容创建页面）
 
（→‎Daily Report）
 
（3位用户的19个中间修订版本未显示）
第6行： 第6行：
 
|-
 
|-
 
| rowspan="3"|2016.10.17
 
| rowspan="3"|2016.10.17
−
|Ying Shi ||   ||     ||     ||
+
|Ying Shi || 9:20am  || 9:00pm  || 11h40m  || DNN and TDNN (th30 wsj)
 
|-
 
|-
−
|Hang luo ||   ||     ||   ||
+
|Hang luo || 9:30am  || 8:20pm  || 10h50m  || paper about joint learning
 
|-
 
|-
−
|Yixiang Chen ||   ||   || ||
+
|Yixiang Chen || 9:30am  || 8:10pm  || 10h40m || channel detection text
 
|-
 
|-
  
第16行： 第16行：
 
|-
 
|-
 
| rowspan="3"|2016.10.18   
 
| rowspan="3"|2016.10.18   
−
|Ying Shi ||   ||     ||     ||   
+
|Ying Shi || 9:20am || 9:20pm || 12h  || DNN and TDNN (wsj) paper  
 
|-
 
|-
−
|Hang luo ||   ||     ||   ||
+
|Hang luo || 9:30am || 7:30pm || 10h  || Read code of joint learning
 
|-
 
|-
−
|Yixiang Chen ||   ||   ||  ||
+
|Yixiang Chen || 9:30am || 7:30pm ||  10h  || channel detection text
 
|-
 
|-
  
  
 +
|-
 +
| rowspan="3"|2016.10.19 
 +
|Ying Shi || 9:30am || 9:00pm  ||  11h30min  ||  visualization job about DNN model
 +
|-
 +
|Hang luo || 9:30am ||  9:30pm ||  12h  || Run original experiment
 +
|-
 +
|Yixiang Chen ||9:30am ||  8:00pm ||  10.5h  || read 'Speakers In The Wild (SITW): The QUT Speaker Recognition System'
 +
|-
  
  
−
|}
+
|-
 +
| rowspan="3"|2016.10.20 
 +
|Ying Shi || 9:30am || 9:00pm  ||  11h30min  ||  visualization job about TDNN model 
 +
|-
 +
|Hang luo || 9:30am ||  7:10pm ||  9h40min  || Modify experiment configurations
 +
|-
 +
|Yixiang Chen ||9:30am ||  8:00pm ||  10.5h  || read 'The Speakers in the Wild (SITW) Speaker Recognition Database'
 +
|-
  
  
 +
|-
 +
| rowspan="3"|2016.10.21 
 +
|Ying Shi || 9:20 || 9:00  || 11h40min  || visualization job about TDNN model 
 +
|-
 +
|Hang luo || 9:20am  || 5:00pm  || 7h40min || Modify configurations and take a meeting
 +
|-
 +
|Yixiang Chen ||9:30am ||  8:30pm ||  11h  || channel detection text
 +
|-
  
  
 +
|-
 +
| rowspan="3"|2016.10.22 
 +
|Ying Shi || 9:20 || 9:00  || 11h40min  || CNN baseline (th30)
 +
|-
 +
|Hang luo ||  ||  ||  ||
 +
|-
 +
|Yixiang Chen || 10:30am  ||  8:00pm  ||  9.5h  || Speaker Recognition Baseline
 +
|-
  
 +
|-
 +
| rowspan="3"|2016.10.23 
 +
|Ying Shi ||  ||  ||  || 
 +
|-
 +
|Hang luo ||  ||  ||  ||
 +
|-
 +
|Yixiang Chen || 11:00am  ||  8:00pm  ||  9h  || Speaker Recognition Baseline debug
 +
|-
 +
 +
|-
 +
| rowspan="3"|2016.10.24 
 +
|Ying Shi || 9:20  || 9:00  || 11h40min  ||  paper reading
 +
|-
 +
|Hang luo ||  ||  ||  ||
 +
|-
 +
|Yixiang Chen || 9:30am  ||  8:00pm  ||  10h30m  || ML-book run Speaker Recognition Baseline |-
 +
 +
 +
|-
 +
| rowspan="3"|2016.10.25 
 +
|Ying Shi || 9:20  || 9:30  || 12h10min  || ML-book (junction tree and some figure) 
 +
|-
 +
|Hang luo ||  ||  ||  ||
 +
|-
 +
|Yixiang Chen || 9:30am  ||  8:00pm  ||  10h30m  || ML-book
 +
|-
 +
 +
|-
 +
| rowspan="3"|2016.10.26 
 +
|Ying Shi || 9:40  || 9:40  || 12h  || CNN (wsj but failed) 
 +
|-
 +
|Hang luo ||  9:30am || 7;30pm  || 10h || Run experiments
 +
|-
 +
|Yixiang Chen || 9:30am  ||  5:30pm  ||  8h  || ML-book and  learing LSTM
 +
|-
 +
 +
|-
 +
| rowspan="3"|2016.10.27 
 +
|Ying Shi || 9:40  || 9:40  || 12h  || visualization job about DNN model   
 +
|-
 +
|Hang luo ||  9:30am || 9:30pm  || 12h || Read paper about Microsoft's research
 +
|-
 +
|Yixiang Chen || 9:30am  ||  7:30pm  ||  10h  || read DNN then LSTM NET code 
 +
|-
 +
 +
 +
|-
 +
| rowspan="3"|2016.10.28 
 +
|Ying Shi || 9:30  || 9:30  || 12h  || plot figure about probability distribution 
 +
|-
 +
|Hang luo || 9:30am  || 5:30pm  || 8h || Read kaldi tutorial
 +
|-
 +
|Yixiang Chen || 9:30am  ||  7:00pm  ||  9h30m  || DNN then LSTM NET
 +
|-
 +
 +
|-
 +
| rowspan="3"|2016.10.29 
 +
|Ying Shi ||  12:00 || 9:40  || 9h40min  ||  make a new model about DNN+lstm
 +
|-
 +
|Hang luo ||  ||  ||  ||
 +
|-
 +
|Yixiang Chen ||  ||    ||      ||
 +
|-
 +
 +
|-
 +
| rowspan="3"|2016.10.30 
 +
|Ying Shi || 12:00  || 9:300pm  || 9h30min  || check the new model DNN+lstm 
 +
|-
 +
|Hang luo || 12:00am  || 5:00om  || 5h || Run experiments
 +
|-
 +
|Yixiang Chen ||  ||    ||      ||
 +
|-
 +
 +
 +
 +
|-
 +
| rowspan="3"|2016.10.31 
 +
|Ying Shi || 10:20am  || 9:30  || 11h30min  || ML-book writing 
 +
|-
 +
|Hang luo ||  ||  ||  ||
 +
|-
 +
|Yixiang Chen ||  ||    ||      ||
 +
|-
 +
 +
|}
  
 
===Time Off Table===
 
===Time Off Table===
第37行： 第153行：
 
! Name      !! Days off
 
! Name      !! Days off
 
|-
 
|-
−
|Ying Shi    || 0.5+1
+
|Ying Shi    || 0.5+1+1
 
|-
 
|-
−
|Hang luo    || from 10.14
+
|Hang luo    || +2+1 ( start from 10.14 )
 
|-
 
|-
−
|Yixiang Chen||  -
+
|Yixiang Chen||  0.5
 
|-
 
|-
  
 
|}
 
|}

2016年11月2日 (三) 01:17的最后版本

Daily Report

Date Person start leave hours status
2016.10.17 Ying Shi 9:20am 9:00pm 11h40m DNN and TDNN (th30 wsj)
Hang luo 9:30am 8:20pm 10h50m paper about joint learning
Yixiang Chen 9:30am 8:10pm 10h40m channel detection text
2016.10.18 Ying Shi 9:20am 9:20pm 12h DNN and TDNN (wsj) paper
Hang luo 9:30am 7:30pm 10h Read code of joint learning
Yixiang Chen 9:30am 7:30pm 10h channel detection text
2016.10.19 Ying Shi 9:30am 9:00pm 11h30min visualization job about DNN model
Hang luo 9:30am 9:30pm 12h Run original experiment
Yixiang Chen 9:30am 8:00pm 10.5h read 'Speakers In The Wild (SITW): The QUT Speaker Recognition System'
2016.10.20 Ying Shi 9:30am 9:00pm 11h30min visualization job about TDNN model
Hang luo 9:30am 7:10pm 9h40min Modify experiment configurations
Yixiang Chen 9:30am 8:00pm 10.5h read 'The Speakers in the Wild (SITW) Speaker Recognition Database'
2016.10.21 Ying Shi 9:20 9:00 11h40min visualization job about TDNN model
Hang luo 9:20am 5:00pm 7h40min Modify configurations and take a meeting
Yixiang Chen 9:30am 8:30pm 11h channel detection text
2016.10.22 Ying Shi 9:20 9:00 11h40min CNN baseline (th30)
Hang luo
Yixiang Chen 10:30am 8:00pm 9.5h Speaker Recognition Baseline
2016.10.23 Ying Shi
Hang luo
Yixiang Chen 11:00am 8:00pm 9h Speaker Recognition Baseline debug
2016.10.24 Ying Shi 9:20 9:00 11h40min paper reading
Hang luo
Yixiang Chen 9:30am 8:00pm 10h30m -


2016.10.25 Ying Shi 9:20 9:30 12h10min ML-book (junction tree and some figure)
Hang luo
Yixiang Chen 9:30am 8:00pm 10h30m ML-book
2016.10.26 Ying Shi 9:40 9:40 12h CNN (wsj but failed)
Hang luo 9:30am 7;30pm 10h Run experiments
Yixiang Chen 9:30am 5:30pm 8h ML-book and learing LSTM
2016.10.27 Ying Shi 9:40 9:40 12h visualization job about DNN model
Hang luo 9:30am 9:30pm 12h Read paper about Microsoft's research
Yixiang Chen 9:30am 7:30pm 10h read DNN then LSTM NET code
2016.10.28 Ying Shi 9:30 9:30 12h plot figure about probability distribution
Hang luo 9:30am 5:30pm 8h Read kaldi tutorial
Yixiang Chen 9:30am 7:00pm 9h30m DNN then LSTM NET
2016.10.29 Ying Shi 12:00 9:40 9h40min make a new model about DNN+lstm
Hang luo
Yixiang Chen
2016.10.30 Ying Shi 12:00 9:300pm 9h30min check the new model DNN+lstm
Hang luo 12:00am 5:00om 5h Run experiments
Yixiang Chen
2016.10.31 Ying Shi 10:20am 9:30 11h30min ML-book writing
Hang luo
Yixiang Chen

Time Off Table

Name Days off
Ying Shi 0.5+1+1
Hang luo +2+1 ( start from 10.14 )
Yixiang Chen 0.5
取自“http://index.cslt.org/mediawiki/index.php?title=Asr-progress_2016.10&oldid=23330”