“2025-02-24”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
 
(3位用户的3个中间修订版本未显示)
第6行: 第6行:
 
|Dong Wang
 
|Dong Wang
 
||
 
||
*  
+
* Live broadcast in Xuetang online
 +
* High-school slides for AIGE check
 +
* Discussion with a couple of people about AI education
  
 
||
 
||
第186行: 第188行:
 
|Yang Wei
 
|Yang Wei
 
||
 
||
*  
+
* Finetune text enroll kws model with accent data with additional fnn.
 
||
 
||
 
*
 
*
第197行: 第199行:
 
||
 
||
 
* Added Language model to ASR
 
* Added Language model to ASR
** Evaluation with LM shows improvement [](https://z1et6d3xtb.feishu.cn/docx/JvDsd8zR4oMwnyxQEQdckpMjn7m?from=from_copylink)
+
** Evaluation with LM shows improvement [https://z1et6d3xtb.feishu.cn/docx/JvDsd8zR4oMwnyxQEQdckpMjn7m?from=from_copylink]
 
* Thesis writing
 
* Thesis writing
 
||
 
||
第215行: 第217行:
 
|Qi Qu
 
|Qi Qu
 
||
 
||
*  
+
* KWS:
 +
** Text-enroll model refactored into 3 sub-models, i.e. `keyword`, `speech` and `score`; [https://b30lttjm7l.feishu.cn/docx/WORcdiE1io86Agxg9hOcKIO0nwe metrics] updated.
 +
** NPU (mr536) demo implemented based on the new 3-component architecture.
 
||
 
||
 
*
 
*

2025年2月24日 (一) 11:59的最后版本

People This Week Next Week Task Tracking (DeadLine)
Dong Wang
  • Live broadcast in Xuetang online
  • High-school slides for AIGE check
  • Discussion with a couple of people about AI education
Lantian Li
Ying Shi
  • Conditional-Chain ASR
    • Design frame level condition fusion method(The model is in training.)
Zhenghai You
Junming Yuan
Xiaolou Li
Zehua Liu
Pengqi Li
Wan Lin
Tianhao Wang
Xiaoxue Luo
Zhenyu Zhou
Junhui Chen
Jiaying Wang
Yu Zhang
Wenqiang Du
  • Primary handbook's PPT (17+4/44)
    • Related xuean(17/44)
  • Investigate cross lingual voice cloning
Yang Wei
  • Finetune text enroll kws model with accent data with additional fnn.
Turi
  • Added Language model to ASR
    • Evaluation with LM shows improvement [1]
  • Thesis writing
Yue Gu
Qi Qu
  • KWS:
    • Text-enroll model refactored into 3 sub-models, i.e. `keyword`, `speech` and `score`; metrics updated.
    • NPU (mr536) demo implemented based on the new 3-component architecture.