“2021-07-05”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
 
(6位用户的9个中间修订版本未显示)
第17行: 第17行:
 
|Yunqi Cai
 
|Yunqi Cai
 
||   
 
||   
*
+
*thermal face recognition dataset & baseline and paper review
 
||
 
||
 
*  
 
*  
第28行: 第28行:
 
|Lantian Li
 
|Lantian Li
 
||  
 
||  
*  
+
* Complete Huawei projects delivery.
 +
* Test hard trials on SOTA models.
 +
||
 +
* Test speech impaired on whisper.
 
||
 
||
 
*  
 
*  
||
 
 
 
|-
 
|-
  
第52行: 第53行:
 
|Haoran Sun
 
|Haoran Sun
 
||  
 
||  
*  
+
* presentation for information bottleneck
 +
* baseline for emotion recognition
 +
* removing content made SER much worse(0.577->0.3201)
 +
* SER gets worse as speechflow training(0.4129->0.3209->0.3201)
 
||
 
||
*  
+
* relationship between ASR and SER
 +
* remove content or just content embedding
 +
* influence of other factors
 +
* try to disentangle emotion by speechflow
 
||
 
||
 
*   
 
*   
第64行: 第71行:
 
|Jiao Han
 
|Jiao Han
 
||  
 
||  
*  
+
* Paper writing about MG exps for APSIPA.
 
||
 
||
*  
+
* Complete the paper submission.
 
||
 
||
 
*   
 
*   
|-
 
 
 
|-
 
|Yang Zhang
 
||
 
 
||
 
*
 
||
 
*
 
 
|-
 
|-
  
第97行: 第93行:
 
|Tiankai Zhi
 
|Tiankai Zhi
 
||  
 
||  
*  
+
* Mongolian ASR baseline ,and paper.
 
||  
 
||  
 
*  
 
*  
第118行: 第114行:
 
|Jingxin Shen
 
|Jingxin Shen
 
||   
 
||   
*  
+
* Collect infrared data sets
 
||
 
||
*  
+
* reproduce SpeakingFaces code
 +
||
 +
 +
|-
 +
 
 +
|-
 +
|Pengqi Li
 +
|| 
 +
*Review work and code of DRT
 +
*Some experiment of DRT
 +
||
 +
*more experiment of DRT
 
||
 
||
 
*   
 
*   

2021年7月12日 (一) 03:27的最后版本

People This Week Next Week Task Tracking (DeadLine)
Dong Wang
  • Review paper for thermal face recognition
  • Continue the thermal review paper
Yunqi Cai
  • thermal face recognition dataset & baseline and paper review
Lantian Li
  • Complete Huawei projects delivery.
  • Test hard trials on SOTA models.
  • Test speech impaired on whisper.
Ying Shi
  • m2asr database
  • mongolian asr baseline
  • mongolian asr baseline
  • zeus
Haoran Sun
  • presentation for information bottleneck
  • baseline for emotion recognition
  • removing content made SER much worse(0.577->0.3201)
  • SER gets worse as speechflow training(0.4129->0.3209->0.3201)
  • relationship between ASR and SER
  • remove content or just content embedding
  • influence of other factors
  • try to disentangle emotion by speechflow
Jiao Han
  • Paper writing about MG exps for APSIPA.
  • Complete the paper submission.
Di Wang
  • The paper of dplda was submitted.
  • Submit the "hard trials" paper on the APSIPA.
Tiankai Zhi
  • Mongolian ASR baseline ,and paper.
Chen Chen
  • Organize speech enhancement backbone(reading paper)
  • Prepare the report about overview of speech enhancement
  • Organize speech enhancement backbone(review & run code)
Jingxin Shen
  • Collect infrared data sets
  • reproduce SpeakingFaces code
Pengqi Li
  • Review work and code of DRT
  • Some experiment of DRT
  • more experiment of DRT