“2021-03-01”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6行: 第6行:
 
|Dong Wang
 
|Dong Wang
 
||  
 
||  
*  
+
* NSFC proposal
 
||
 
||
*  
+
* NSFC proposal
 
||
 
||
 
*   
 
*   
第17行: 第17行:
 
|Yunqi Cai
 
|Yunqi Cai
 
||   
 
||   
*  
+
* simulation experiments of uniform distribution
 +
* IR investigation
 +
* NCSF founding proposal
 
||
 
||
 
*  
 
*  
第28行: 第30行:
 
|Lantian Li
 
|Lantian Li
 
||  
 
||  
*  
+
* Performance measurement for ASV evaluation.
 +
* Compare Softmax, L-Softmax, AM-Softmax and ARM-Softmax in THUCS30 (Ruiqian).
 +
* Glance Sunine, still lots of problems.
 
||
 
||
*  
+
* Verify ARM-Softmax in VoxCeleb.
 +
* Quantitive analysis on DRT.
 
||
 
||
 
*   
 
*   
第39行: 第44行:
 
|Ying Shi
 
|Ying Shi
 
||  
 
||  
*  
+
* analysis DAE-iVector, DAE-NAP
 
||
 
||
*  
+
* continue on DAE-iVector
 
||
 
||
 
*   
 
*   
第50行: 第55行:
 
|Haoran Sun
 
|Haoran Sun
 
||  
 
||  
*  
+
* continuous i-vector
 
||
 
||
 
*  
 
*  
第85行: 第90行:
 
|Di Wang
 
|Di Wang
 
||  
 
||  
*  
+
* Experiments to construct different trials.
 
||
 
||
*  
+
* Analyze the test scheme in speaker evaluation.
 
||
 
||
 
*   
 
*   

2021年3月8日 (一) 01:29的最后版本

People This Week Next Week Task Tracking (DeadLine)
Dong Wang
  • NSFC proposal
  • NSFC proposal
Yunqi Cai
  • simulation experiments of uniform distribution
  • IR investigation
  • NCSF founding proposal
Lantian Li
  • Performance measurement for ASV evaluation.
  • Compare Softmax, L-Softmax, AM-Softmax and ARM-Softmax in THUCS30 (Ruiqian).
  • Glance Sunine, still lots of problems.
  • Verify ARM-Softmax in VoxCeleb.
  • Quantitive analysis on DRT.
Ying Shi
  • analysis DAE-iVector, DAE-NAP
  • continue on DAE-iVector
Haoran Sun
  • continuous i-vector
Jiao Han
Yang Zhang
  • Read some paper about model compression and inference acceleration.
  • finish VoxCebel and CN-Celeb dataset code usage examples, and SITW training part.
  • train more model, and upload the checkpoint to CSLT wiki.
Di Wang
  • Experiments to construct different trials.
  • Analyze the test scheme in speaker evaluation.
Tiankai Zhi