“2023-09-04”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
 
(5位用户的5个中间修订版本未显示)
第30行: 第30行:
 
* information agent token for my  Overlap asr methods
 
* information agent token for my  Overlap asr methods
 
* training model with  librispeech 960h
 
* training model with  librispeech 960h
* writer icassp 2024 paper  
+
* writer icassp 2024 paper [https://z1et6d3xtb.feishu.cn/docx/LDRydgHmeoWlM6x23wfcGHCwn5c here]
 
||
 
||
 
* do more test
 
* do more test
第42行: 第42行:
 
|Chen Chen
 
|Chen Chen
 
||  
 
||  
*
+
* check CNVSRC code&data
 +
* check SQUIM [https://z1et6d3xtb.feishu.cn/docx/LnSedYxFcoIGfixczTGcFmeYngf]
 
||
 
||
*
+
* use sync-loss and squim-loss for VTS training
 
||
 
||
 
*   
 
*   
第53行: 第54行:
 
|Pengqi Li
 
|Pengqi Li
 
||   
 
||   
*
+
* Experiment [https://z1et6d3xtb.feishu.cn/docx/NRM3dCsX6onFImxASAycW4Oqnfd]
 
||
 
||
 
*
 
*
第65行: 第66行:
 
|Wan Lin
 
|Wan Lin
 
||  
 
||  
*
+
* Chinese edition of the paper [https://z1et6d3xtb.feishu.cn/docx/QT3kdtajRo9576xLgWPcpcnRnBk?from=from_copylink]
 +
* Clustering domain adaptation exploration [https://z1et6d3xtb.feishu.cn/docx/OnzGdXfnboNLISxh5rIcHh7knuf?from=from_copylink]
 
||
 
||
 
*
 
*
第76行: 第78行:
 
|Tianhao Wang
 
|Tianhao Wang
 
||  
 
||  
*
+
* genre-based SE adapter
 
||
 
||
 
*
 
*
第109行: 第111行:
 
|Xiaolou Li
 
|Xiaolou Li
 
||  
 
||  
*  
+
* CNVSRC baseline[https://z1et6d3xtb.feishu.cn/docx/IOLedrVmNoLOpTxlBwHc37xWnQo]
 +
* Single model infer on CNC-AV[https://z1et6d3xtb.feishu.cn/docx/Z2KYdQGmooNtboxJPJVcKcyOndf] and VoxCeleb1[https://z1et6d3xtb.feishu.cn/docx/YaPRdB685okRRfxQzASc8fBknWh]
 
||
 
||
 
*  
 
*  

2023年9月4日 (一) 11:01的最后版本

People This Week Next Week Task Tracking (DeadLine)
Dong Wang
  • AIGraph publication
  • Revise MG paper
Lantian Li
  • CIPS Youth project
  • ICASSP 2024
Ying Shi
  • BPE token for my Overlap asr methods
  • information agent token for my Overlap asr methods
  • training model with librispeech 960h
  • writer icassp 2024 paper here
  • do more test
  • finish my paper
Chen Chen
  • check CNVSRC code&data
  • check SQUIM [1]
  • use sync-loss and squim-loss for VTS training
Pengqi Li
  • Experiment [2]
Wan Lin
  • Chinese edition of the paper [3]
  • Clustering domain adaptation exploration [4]
Tianhao Wang
  • genre-based SE adapter
Zhenyu Zhou
  • ICASSP 2024
Jiaying Wang
  • recovering
Xiaolou Li
  • CNVSRC baseline[5]
  • Single model infer on CNC-AV[6] and VoxCeleb1[7]
Junhui Chen
  • ICASSP 2024
Wenqiang Du
  • Some project cooperation
Yang Wei
  • Manage some old services