“Task List”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
 
(4位用户的19个中间修订版本未显示)
第1行: 第1行:
=Task To Do=
 
==Speech Recognition==
 
*End-to-End speech recognition
 
:*Zhiyuan Tang/Mengyuan Zhao/Zhiyong Zhang
 
  
*Integrate the class information to HCLG fst for speech recognition
+
=Tasks at hand=
 
+
*Distant speech recognition
+
:*RNN-DAE: echo or reverberation
+
::*Xuewei Zhang/Zhiyuan Tang/Mengyuan Zhao/Zhiyong Zhang
+
:*Reverberation
+
::*Mutli-microphones
+
::*(Lasso),Xuewei Zhang
+
 
+
*Voice conversation
+
 
+
*Unbound activation function(Rectifier/Maxout/Pnorm) go-through searching method
+
:* RNN-Rectifier
+
::*Zhiyuan Tang
+
:* P-norm
+
 
+
*Sparse DNN
+
:*Zhiyuan Tang
+
 
+
*Monmentum-like Hessien-Free acceleration
+
:*Zhiyong Zhang
+
 
+
*Correlation based SENONE cluster
+
 
+
 
+
*NN Multi-GPU parallel traing
+
:*Multi-Machine
+
::* nnet2 optimization
+
::*Sheng Su
+
:*Multi-GPU on one Machine
+
::*Sheng Su
+
 
+
*Audio Embedding
+
:*Ke Ning
+
 
+
*Activation value normalization through time
+
:* For bigger learning rate
+
 
+
*Mix-training Balance decision tree
+
:* Zhiyong Zhang
+
 
+
*RNN training accelerating
+
 
+
*Data selection
+
:*Zhiyong Zhang
+
:*Sub-modular data selection
+
 
+
*Decoder
+
:*Confidence output for task-required
+
 
+
*xx-h Chinese data-set release
+
:Xuewei Zhang
+
  
 +
==Speech Recognition==
  
==Speaker Verification==
+
===joint learning===
*binary code
+
* Hang Luo, Zhiyuan Tang
:*Lantian Li
+
  
*RNN-ivector
+
===visualization===
:*Lantian Li
+
* Ying Shi, Zhiyuan Tang
  
*DNN clustering
+
==Speaker Recognition==
:*Lantian Li
+
*Lantian Li, Yixiang Chen
  
=Task DONE=
 
*Multi-Mode features based VAD
 
:*Shi Yin
 
  
*DNN based Language identification and Speaker identification
+
=Tasks Done=
:*Xuewei Zhang/Zhiyuan Tang
+
  
*Neural network visulization
+
=Technical Reports to write=
:*Mian Wang,DONE
+
  
*Dark knowledge
+
=Papers to write=
:*Mengyuan Zhao, Xiangyu Zeng, Zhiyong Zhang, Chao Liu
+
  
*Normal RNN speech recognition
+
=Patents to write=
:*Mengyuan Zhao
+
  
=Technical Report To Write=
+
=Patents done=
1, DNN-DAE based noise cancellation -- Xiangyu Zeng / Mengyuan Zhao / Zhiyong Zhang  --DONE
+
2, Speech Rate DNN speech recognition --Shi Yin/Xiangyu Zeng --DONE
+
3, CNN+fbank feature combination --Mian Wang /Yiye Lin /Mengyuan Zhao /Shi Yin
+
4, Uyghur low-resource acoustic model enhancement -- Shi Yin / Mengyuan Zhao / Zhiyong Zhang --DONE
+
5, Uyghur 20h database release --Kaer /Shi Yin --DONE
+
6,Dark-Knowledge Transfer
+
    *: Xiangyu Zeng/ Mengyuan Zhao / Zhiyong Zhang
+
  
=Paper to Write=
+
=Projects=
  
=Project=
 
* Xiaomi TV
 
:*Mengyuan Zhao/Zhiyong Zhang
 
:*TAG-lm & Domain-specific general lm
 
  
*Chinese-English mix-training
+
------------------------------
 +
[[task previous]]

2016年10月16日 (日) 12:31的最后版本

Tasks at hand

Speech Recognition

joint learning

  • Hang Luo, Zhiyuan Tang

visualization

  • Ying Shi, Zhiyuan Tang

Speaker Recognition

  • Lantian Li, Yixiang Chen


Tasks Done

Technical Reports to write

Papers to write

Patents to write

Patents done

Projects


task previous