“ASR Status Report 2017-5-22”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
 
(2位用户的2个中间修订版本未显示)
第27行: 第27行:
 
|Ying Shi   
 
|Ying Shi   
 
||  
 
||  
*  
+
* work for nips 
 +
* condition training done (bad performance)
 +
* gate control for joint system
 
||  
 
||  
*  
+
* work for apsipa17
 
|-
 
|-
  
第36行: 第38行:
 
|Yixiang Chen   
 
|Yixiang Chen   
 
||  
 
||  
*   
+
* work for nips  
 
||  
 
||  
 
*  
 
*  
第45行: 第47行:
 
|Lantian Li   
 
|Lantian Li   
 
||  
 
||  
*  
+
* Deep Factorization for Speech Signal.
 
||  
 
||  
*  
+
* Deep Factorization for Speech Signal.
 
|-
 
|-
 +
  
  
第117行: 第120行:
 
|Lantian Li   
 
|Lantian Li   
 
||  
 
||  
*  
+
* Deep Factorization for Speech Signal.
 
||  
 
||  
*  
+
* Deep Factorization for Speech Signal.
 
|-
 
|-
 +
  
  

2017年6月1日 (四) 04:51的最后版本

Date People Last Week This Week
2017.5.22


Hui Tang
  • learned the principle of beamforming and the knowledge about signal process.
  • fix beamforming: DS SD,MVDR
  • the adaptive beamforming: LMS...
Yanqing Wang
  • Implemetation of neural sparseness on wsj/thchs30
  • Baseline of wsj and thchs30
Ying Shi
  • work for nips
  • condition training done (bad performance)
  • gate control for joint system
  • work for apsipa17
Yixiang Chen
  • work for nips
Lantian Li
  • Deep Factorization for Speech Signal.
  • Deep Factorization for Speech Signal.
Zhiyuan Tang
  • Material (spoken language processing, DL) reading for Kaldi book writing.
  • Proofreading on NIPS17 paper.
  • MixASR-CHEN 2017 challenge / AP17-OLR challenge.
  • Material reading.

Date People Last Week This Week
2017.5.15


Hui Tang
  • read the code of BTK.
  • to prepare a document or ppt for introducing the ways of beamforming in two weeks.
    • this week: SD, DS, MVDR
    • next week: MMSE, other wasys
  • continue to read the code of BTK
Yanqing Wang
  • basic implementation of sparse neural network on wsj/thchs30
    • baseline
    • neural sparseness
  • a review on sparse neural network
Ying Shi
  • kazak speech recognition
    • less pdf and position feature training down decoding is in progress
    • work for NIPS with Lilt
  • joint system noise decoding here
  • work for NIPS with Lilt
Yixiang Chen
Lantian Li
  • Deep Factorization for Speech Signal.
  • Deep Factorization for Speech Signal.
Zhiyuan Tang
  • ICASSP17 review on selected papers. Schedule for Kaldi book writing.
  • Paper reading with ICASSP17. Book writing.