cslt Wiki - 用户贡献 [zh-cn]

FreeNeb status Report 2019-05-13

2019-05-13T03:50:38Z

Wangd：

'''警告：'''“FreeNeb status Report 2019-05-13”指向这里，但您没有足够的权限来访问它。

FreeNeb Status Report 2017-04-17

2017-04-17T02:18:44Z

Wangd：

'''警告：'''“FreeNeb Status Report 2017-04-17”指向这里，但您没有足够的权限来访问它。

FreeNeb Status Report 2017-04-17

2017-04-17T02:18:15Z

Wangd：

'''警告：'''“FreeNeb Status Report 2017-04-17”指向这里，但您没有足够的权限来访问它。

2016 Summer Seminar for Machine learning

2016-07-08T11:21:47Z

Wangd：

*Location: FIT-1-304

{| class="wikitable"
! Date !! Speaker!! Title !! Materials
|-
| 2016/07/04 ||Dong Wang || Machine learning overview || [http://wangd.cslt.org/talks/seminar/2016-sum-ml/chpt1.%20Overview%20of%20Machine%20Learning.pptx slides][http://cs229.stanford.edu/section/cs229-linalg.pdf Algebra review] [http://cs229.stanford.edu/section/cs229-prob.pdf probability review] [http://cs229.stanford.edu/section/gaussians.pdf Gaussian distribution][http://cs229.stanford.edu/notes/cs229-notes4.pdf Learning theory]
|-
| 2016/07/05 ||Dong Wang || Linear models || [http://wangd.cslt.org/talks/seminar/2016-sum-ml/chpt2.%20Linear%20Model.pptx slides] [http://cs229.stanford.edu/notes/cs229-notes1.pdf NG's lecture 1] [http://cs229.stanford.edu/notes/cs229-notes2.pdf NG's lecture 2]
|-
| 2016/07/ ||Dong Wang || Neural networks || [http://wangd.cslt.org/talks/seminar/2016-sum-ml/chpt3.%20Neural%20Network.pptx slides]
|-
| 2016/07/ ||Dong Wang || Deep learning (1)||
|-
| 2016/07/ ||Dong Wang || Deep learning (2)||
|-
| 2016/07/ ||Dong Wang || Unsupervised learning ||
|-
| 2016/07/ ||Caixia Wang || Kernel methods || [http://cslt.riit.tsinghua.edu.cn/mediawiki/images/b/b5/Kernel_Methods_for_Pattern_Analysis.pdf Kernel method book] [http://cslt.riit.tsinghua.edu.cn/mediawiki/images/9/95/Pattern_Recognition_and_Machine_Learning.pdf pattern recognition 6-7]
|-
| 2016/07/ ||Dong Wang || Probabilistic learning theory||
|-
| 2016/07/ ||Yang Feng || Graphical model: Bayesian approach ||
|-
| 2016/07/ ||Yang Feng || Graphical model: Random field ||
|-
| 2016/07/ ||Dong Wang || No parametric models || [http://cs229.stanford.edu/section/cs229-gaussian_processes.pdf Gaussian process]
|-
| 2016/07/ ||Dong Wang || Reinforcement learning ||
|-
| 2016/07/ ||Maoning Wang || Evolutionary learning ||
|-
| 2016/07/ ||Dong Wang || Optimization || [http://cs229.stanford.edu/section/cs229-cvxopt.pdf Convex optimization I] [http://cs229.stanford.edu/section/cs229-cvxopt.pdf Convex optimization II]
|-
|}

2016 Summer Seminar for Machine learning

2016-07-08T09:53:14Z

Wangd：

*Location: FIT-1-304

{| class="wikitable"
! Date !! Speaker!! Title !! Materials
|-
| 2016/07/04 ||Dong Wang || Machine learning overview || [http://wangd.cslt.org/talks/seminar/2016-sum-ml/chpt1.%20Overview%20of%20Machine%20Learning.pptx slides][http://cs229.stanford.edu/section/cs229-linalg.pdf Algebra review] [http://cs229.stanford.edu/section/cs229-prob.pdf probability review] [http://cs229.stanford.edu/section/gaussians.pdf Gaussian distribution][http://cs229.stanford.edu/notes/cs229-notes4.pdf Learning theory]
|-
| 2016/07/05 ||Dong Wang || Linear models || [http://wangd.cslt.org/talks/seminar/2016-sum-ml/chpt2.%20Linear%20Model.pptx slides] [http://cs229.stanford.edu/notes/cs229-notes1.pdf NG's lecture 1] [http://cs229.stanford.edu/notes/cs229-notes2.pdf NG's lecture 2]
|-
| 2016/07/ ||Dong Wang || Neural networks || [http://wangd.cslt.org/talks/seminar/2016-sum-ml/chpt2.%20Neural%20Network.pptx slides]
|-
| 2016/07/ ||Dong Wang || Deep learning (1)||
|-
| 2016/07/ ||Dong Wang || Deep learning (2)||
|-
| 2016/07/ ||Dong Wang || Unsupervised learning ||
|-
| 2016/07/ ||Caixia Wang || Kernel methods || [http://cslt.riit.tsinghua.edu.cn/mediawiki/images/b/b5/Kernel_Methods_for_Pattern_Analysis.pdf Kernel method book] [http://cslt.riit.tsinghua.edu.cn/mediawiki/images/9/95/Pattern_Recognition_and_Machine_Learning.pdf pattern recognition 6-7]
|-
| 2016/07/ ||Dong Wang || Probabilistic learning theory||
|-
| 2016/07/ ||Yang Feng || Graphical model: Bayesian approach ||
|-
| 2016/07/ ||Yang Feng || Graphical model: Random field ||
|-
| 2016/07/ ||Dong Wang || No parametric models || [http://cs229.stanford.edu/section/cs229-gaussian_processes.pdf Gaussian process]
|-
| 2016/07/ ||Dong Wang || Reinforcement learning ||
|-
| 2016/07/ ||Maoning Wang || Evolutionary learning ||
|-
| 2016/07/ ||Dong Wang || Optimization || [http://cs229.stanford.edu/section/cs229-cvxopt.pdf Convex optimization I] [http://cs229.stanford.edu/section/cs229-cvxopt.pdf Convex optimization II]
|-
|}

Xingsong-meeting-minutues-2016-06-15

2016-06-15T01:47:05Z

Wangd：

==项目进展==

===语音识别系统===

1. 数据准备

#第四次采集完成

2. 基线系统

#基线识别系统完成

3. 麦克风阵列研究 ([[xinsong-beamforming-result|测试结果页]])

*缺省beamforming

*Beamforming算法研究
:SD 方法包装与工程化

*Adaptation算法优化
:single layer adaptation

* Best system
* 8.24%

4. 模型适应

#LM tag模型完成

===语义理解系统===

#语义理解模块包装
#语音容错对应处理
#对话完成率:文字输入 98%
#对话完成率:语音输入 90%

===系统实现===

#系统集成
#Android client开发

==6月份工作计划==

===系统集成===
* NLP 系统联调一周内完成
* Android与系统联调，一周内完成

Xingsong-meeting-minutues-2016-06-15

2016-06-15T01:13:09Z

Wangd：以“==项目进展== ===语音识别系统=== 1. 数据准备 #第四次采集完成 2. 基线系统 #基线识别系统完成 3. 麦克风阵列研究 (xinson...”为内容创建页面

==项目进展==

===语音识别系统===

1. 数据准备

#第四次采集完成

2. 基线系统

#基线识别系统完成

3. 麦克风阵列研究 ([[xinsong-beamforming-result|测试结果页]])

*缺省beamforming

*Beamforming算法研究
:SD 方法包装与工程化

*Adaptation算法优化
:single layer adaptation

4. 模型适应

#LM tag模型完成

===语义理解系统===

#语义理解模块包装
#语音容错对应处理开

===系统实现===

#系统集成
#Android client开发

==5工作计划==

===ASR系统性能优化(CSLT)===

*beamforming数据采集
*noisy training
*dark knowledge training
*multi-channel DNN training
*similar-pair LM

===基于ASR的NLP实验===
*example output generation (Sion + CSLT),before 5/18
*ASR-oriented NLP (CS)

===系统集成===
*prototype system with NLP (Sino+CS)
:* NLP submodule ready before 5/30
:* NLP output structure deliver before 5/16
:* Xinsong develop Android application 6/10

Xinsong-doc

2016-06-15T00:52:56Z

Wangd：

[[xingsong-project-plan|技术方案]]

[[xingsong-meeting-minutues-2016-04-6|会议记录(2016-04-06)]]

[[xingsong-meeting-minutues-2016-05-12|会议记录(2016-05-12)]]

[[xingsong-meeting-minutues-2016-06-15|会议记录(2016-06-15)]]

Xingsong-meeting-minutues-2016-05-12

2016-05-12T02:05:51Z

Wangd：

==项目进展==

===语音识别系统===

1. 数据准备

#麦克风阵列采集完成三轮
#标注与切分方法确定
#继续采集30-50小时数据

2. 基线系统

#基线识别系统完成

3. 麦克风阵列研究[http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=wangd&step=view_request&cvssid=529][http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=wangd&step=view_request&cvssid=535]

*缺省beamforming

*Beamforming算法研究
:*delay & sum (DS)
:*super direction (SD)
:*MVDR

*DAE算法
:*single channel
:*multi channel
:*BM+DAE

4. 模型适应

#字模型+领域相关词模型测试完成，识别结果与大规模词模型相似

===语义理解系统===

#语义理解，对话管理模块bug fix
#语音容错对应处理开始

===系统实现===

#系统主要模块完成
#等待NLP和Android client模块集成

==5月工作计划==

===ASR系统性能优化(CSLT)===

*beamforming数据采集
*noisy training
*dark knowledge training
*multi-channel DNN training
*similar-pair LM

===基于ASR的NLP实验===
*example output generation (Sion + CSLT),before 5/18
*ASR-oriented NLP (CS)

===系统集成===
*prototype system with NLP (Sino+CS)
:* NLP submodule ready before 5/30
:* NLP output structure deliver before 5/16
:* Xinsong develop Android application 6/10

Xingsong-meeting-minutues-2016-04-6

2016-04-06T01:42:08Z

Wangd：

==项目进展==

===语音识别系统===

1. 数据准备

(1) clean 数据采集完成， noisy手机数据采集100句
(2) 麦克风阵列采集开始准备

2. 基线系统

(1) clean数据测试完成, noisy手机数据测试完成。

3. 麦克风阵列研究

4. 模型适应

(1) LM自适应完成。

===语义理解系统===

1. 数据收集与梳理

2. 语义理解，对话管理模块完成。

3. 语音容错对应处理开始

===系统实现===

1. 系统移植完成

2. 模型裁剪未验证

3. 接口设计完成

==工作计划==

1. ASR/NLP 接口数据

(1) n-best word (2) n-best phone (3) phone-based confidence (4) syllable-based confidence (5) word-based confidence

2. mic-array数据录制

(1) 录音方案设计（捷通）
(2) 录音（清华，新松）

Xingsong-meeting-minutues-2016-04-6

2016-04-06T00:51:44Z

Wangd：以“==项目进展== ===语音识别系统=== 1. 数据准备 2. 基线系统 3. 麦克风阵列研究 4. 模型适应 ===语义理解系统=== 1, 数据收集与梳...”为内容创建页面

==项目进展==

===语音识别系统===
1. 数据准备
2. 基线系统
3. 麦克风阵列研究
4. 模型适应

===语义理解系统===
1, 数据收集与梳理
2. 模型训练
3.

===系统实现===
1. 系统移植
2. 模型裁剪
3. 接口设计

==工作计划==

Xinsong-doc

2016-04-06T00:46:37Z

Wangd：

[[xingsong-project-plan|技术方案]]

[[xingsong-meeting-minutues-2016-04-6|会议记录(2016-04-06)]]

Wangd-kaldi-book

2016-02-11T09:49:43Z

Wangd：

'''警告：'''“Wangd-kaldi-book”指向这里，但您没有足够的权限来访问它。

Wangd-kaldi-book

2016-02-11T09:44:38Z

Wangd：以“=Kaldi从零开始：当代语音识别系统设计与开发= ==绪论（王东）== ===研究历史与现状=== ===研究方法=== ===语音识别应用现状与...”为内容创建页面

'''警告：'''“Wangd-kaldi-book”指向这里，但您没有足够的权限来访问它。

Dongwang-personal

2016-02-11T09:02:57Z

Wangd：

[[wangd-apsipa2015]]

[[wangd-kaldi-book]]

Dongwang-personal

2016-02-11T09:02:47Z

Wangd：

[[wangd-apsipa2015]]
[[wangd-kaldi-book]]

Wangd-apsipa2015

2016-02-11T08:56:10Z

Wangd：以“APSIPA 05: Chair/Secretary/past chair + committee = 3 + 38 = 41 China: 8 Japan:7 Sigapore: 4 Taiwan: 4 USA: 4 Australia:3 Korea:3 HK: 2 Vienem: 2 Thailand: 1 New...”为内容创建页面

APSIPA 05:

Chair/Secretary/past chair + committee = 3 + 38 = 41

China: 8
Japan:7
Sigapore: 4
Taiwan: 4
USA: 4
Australia:3
Korea:3
HK: 2
Vienem: 2
Thailand: 1
Newzeeland: 1
Canada: 1
India: 1

* According to by law, we can have another two members.
* Better to involve more diversity

*Woon-Seng Gan (2015-2016) (Dec/2016*), Nanyang Technological University (Singapore)
*Bin Ma (Dec/2017*), Institute for Infocomm Research (Singapore)
*Susanto Rahardja (Dec/2016**), Institute for Infocomm Research (Singapore)
*Rongshan Yu (Dec/2016*), Institute for Infocomm Research (Singapore)

*Jianwu Dang (Dec/2017**), Tianjin Univ. (China)
*Dong Wang (2015-2016) (Dec/2016*), Tsinghua University (China)
*Askar Hamdulla (Dec/2016*), Xinjiang University of China (China)
*Qin Jin (Dec/2017*), Renmin University of China (China)
*Frank Soong (Dec/2016**), Microsoft Research (China)
*Jianhua Tao (Dec/2016**), CAS - Institute of Automation (China)
*Lei Xie (Dec/2017**), Northwestern Polytech Univ. (China)
*Thomas Zheng (Dec/2016**), Tsinghua Univ. (China)

*Chung-Hsien Wu (Dec/2014), National Cheng Kung University (Taiwan)
*Jen-Tzung Chien (Dec/2016**), National Chiao Tung University (Taiwan)
*Lin-shan Lee (Dec/2017**), National Taiwan University (Taiwan)
*Hsin-Min Wang (Dec/2016**), Academia Sinica (Taiwan)

*Eliathamby Ambikairajah (Dec/2017**), Univ. New South Wales (Australia)
*Christian Ritz (Dec/2017*), University of Wollongong (Australia)
*Michael Wagner (Dec/2016**), Univ. Canberra (Australia)

*Wutiwiwatchai Chai (Dec/2017**), NECTEC (Thailand)

*Pascale Fung (Dec/2016**), HKUST (Hong Kong)
*Helen Meng (Dec/2016**), CUHK (Hong Kong)

*Masataka Goto (Dec/2016**), AIST (Japan)
*Yoshinobu Kajikawa (Dec/2017*), Kansai University (Japan)
*Tatsuya Kawahara (Dec/2016**), Kyoto University (Japan)
*Shoji Makino (Dec/2016**), Univ. Tsukuba (Japan)
*Hiroshi Sawada (Dec/2016**), NTT (Japan)
*Koichi Shinoda (Dec/2016*), Tokyo Institute of Technology (Japan)
*Tomoki Toda (Dec/2016**), NAIST (Japan)

*Mark Hasegawa-Johnson (Dec/2016**), Univ. Illinois UC (USA)
*Chin-Hui Lee (Dec/2016**), Georgia Tech (USA)
*Shinji Watanabe (Dec/2016*), Mitsubishi Electric Research Laboratories (USA)
*Li Deng (Dec/2016**), Microsoft Research (USA)

*Hui Jiang (Dec/2016**), York Univ. (Canada)

*Hoirin Kim (Dec/2016**), KAIST (Korea)
*Nam-Soo Kim (Dec/2016**), Seoul National Univ. (Korea)
*Gary Geunbae Lee (Dec/2016**), POSTECH (Korea)

*Luong Chi Mai (Dec/2017**), Vietnamese Academy of Science and Technology (Vietnam)
*Vu Hai Quan (Dec/2017**), Vietnam National Univ. (Vietnam)

*Srinivasan Umesh (Dec/2016*), Indian Institute of Technology-Madras (India)

*Abdulla Waleed (Dec/2016**), Univ. Auckland (New Zealand)

Dongwang-personal

2016-02-11T08:56:00Z

Wangd：以“wangd-apsipa2015”替换内容

[[wangd-apsipa2015]]

Dongwang-personal

2015-12-16T14:30:45Z

Wangd：

Dongwang-personal

2015-12-16T14:23:42Z

APSIPA 05:

Chair/Secretary/past chair + committee = 3 + 38 = 41

China: 8
Japan:7
Sigapore: 4
Taiwan: 4
USA: 4
Australia:3
Korea:3
HK: 2
Vienem: 2
Thailand: 1
Newzeeland: 1
Canada: 1
India: 1

* According to by law, we can have another two members.
* Better to involve more diversity

Woon-Seng Gan (2015-2016) (Dec/2016*), Nanyang Technological University (Singapore)
Bin Ma (Dec/2017*), Institute for Infocomm Research (Singapore)
Susanto Rahardja (Dec/2016**), Institute for Infocomm Research (Singapore)
Rongshan Yu (Dec/2016*), Institute for Infocomm Research (Singapore)

Jianwu Dang (Dec/2017**), Tianjin Univ. (China)
Dong Wang (2015-2016) (Dec/2016*), Tsinghua University (China)
Askar Hamdulla (Dec/2016*), Xinjiang University of China (China)
Qin Jin (Dec/2017*), Renmin University of China (China)
Frank Soong (Dec/2016**), Microsoft Research (China)
Jianhua Tao (Dec/2016**), CAS - Institute of Automation (China)
Lei Xie (Dec/2017**), Northwestern Polytech Univ. (China)
Thomas Zheng (Dec/2016**), Tsinghua Univ. (China)

Chung-Hsien Wu (Dec/2014), National Cheng Kung University (Taiwan)
Jen-Tzung Chien (Dec/2016**), National Chiao Tung University (Taiwan)
Lin-shan Lee (Dec/2017**), National Taiwan University (Taiwan)
Hsin-Min Wang (Dec/2016**), Academia Sinica (Taiwan)

Eliathamby Ambikairajah (Dec/2017**), Univ. New South Wales (Australia)
Christian Ritz (Dec/2017*), University of Wollongong (Australia)
Michael Wagner (Dec/2016**), Univ. Canberra (Australia)

Wutiwiwatchai Chai (Dec/2017**), NECTEC (Thailand)

Pascale Fung (Dec/2016**), HKUST (Hong Kong)
Helen Meng (Dec/2016**), CUHK (Hong Kong)

Masataka Goto (Dec/2016**), AIST (Japan)
Yoshinobu Kajikawa (Dec/2017*), Kansai University (Japan)
Tatsuya Kawahara (Dec/2016**), Kyoto University (Japan)
Shoji Makino (Dec/2016**), Univ. Tsukuba (Japan)
Hiroshi Sawada (Dec/2016**), NTT (Japan)
Koichi Shinoda (Dec/2016*), Tokyo Institute of Technology (Japan)
Tomoki Toda (Dec/2016**), NAIST (Japan)

Mark Hasegawa-Johnson (Dec/2016**), Univ. Illinois UC (USA)
Chin-Hui Lee (Dec/2016**), Georgia Tech (USA)
Shinji Watanabe (Dec/2016*), Mitsubishi Electric Research Laboratories (USA)
Li Deng (Dec/2016**), Microsoft Research (USA)

Hui Jiang (Dec/2016**), York Univ. (Canada)

Hoirin Kim (Dec/2016**), KAIST (Korea)
Nam-Soo Kim (Dec/2016**), Seoul National Univ. (Korea)
Gary Geunbae Lee (Dec/2016**), POSTECH (Korea)

Luong Chi Mai (Dec/2017**), Vietnamese Academy of Science and Technology (Vietnam)
Vu Hai Quan (Dec/2017**), Vietnam National Univ. (Vietnam)

Srinivasan Umesh (Dec/2016*), Indian Institute of Technology-Madras (India)

Abdulla Waleed (Dec/2016**), Univ. Auckland (New Zealand)

People

2015-12-16T13:52:22Z

Wangd：

'''Faculty'''

Dong Wang [[http://cslt.org:8081/homepages/wangd/ homepage]][Resume [[Dong Wang - resume|(Chinse)]][[Dong Wang - resumeEng|(English)]] ] [ [[wangdong-pic|Pictures]] ][[dongwang-personal|Personal]]

'''Students'''

'''Engineers'''

Xuewei Zhang: group assistant

Zhiyong Zhang: leader of the speech processing team

Chao Xing: Leader of the langauge processing team

Rong Liu : leader of the finance processing team

Mengyuan Zhao

Xiaoxi Wang

[http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/Tianyi_Luo Tianyi Luo]

'''Intern students'''

Zhi Yuan Tang: CAS Chengdu

Dongxu Zhang: BUPT

Xiangyu Zeng: BUPT

Yang Yu: Central University of Fiance and Economics

Hui Tang: University of Science at Chongqing

'''Collaborative students'''

[http://lilt.cslt.org/ Lantian Li]

Askar Rouz: CSLT, VPR group

'''Collaborative researchers'''

Javier Tejedor, University of Madrid, Spain

Nick Evance, EURECOM, France

Simon King, University of Edinburgh, UK

Ravi Vipperla, Nuance, UK

Guozhen Zhao, Chinese Academy

'''Collaborative engineers'''

Xinkai Du, Huilan

Xiaoming Zheng, Sinovoice

'''Allumi'''

Jun Wang: CSLT, VPR group

Sheng Su: BUPT

Ke Ning: Beijing Institute of Technology

Haichao Yu: BUPT

Chao Yuan Zuo: Beijing Institute of Technology

Xi Ma: Beijing Post U.

Xizi Ouyang : Kuang yuan

Mahpirat: Xinjiang U.

Zulhumar, Xinjiang U.

Jun Qi, U. Washington

Hussein: Xinjiang U.

Fuhe Li: Petro Univ. Tsingdao

Tanche Li: CS, Tsinghua Univ.

Xuesheng Li: Petro Univ. Tsingdao

Xiangtao Meng: Chongqing Post U.

Xue Chen, Huilan

Zhao Fang, Huilan

Huabo Xie, Puqiang

Chao Liu

Bin Yuan, BUPT

Shi Yin: Chongqing Post U.

Yiye Lin: Beijing Institute of Technology

Fanhu Bie: CSLT, VPR group

Bo Hu, Puqiang

Mian Wang: BUPT

People

2015-12-16T13:52:00Z

Wangd：

'''Faculty'''

Dong Wang [[http://cslt.org:8081/homepages/wangd/ homepage]][Resume [[Dong Wang - resume|(Chinse)]][[Dong Wang - resumeEng|(English)]] ] [ [[wangdong-pic|Pictures]] ][[Personal]]

'''Students'''

'''Engineers'''

Xuewei Zhang: group assistant

Zhiyong Zhang: leader of the speech processing team

Chao Xing: Leader of the langauge processing team

Rong Liu : leader of the finance processing team

Mengyuan Zhao

Xiaoxi Wang

[http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/Tianyi_Luo Tianyi Luo]

'''Intern students'''

Zhi Yuan Tang: CAS Chengdu

Dongxu Zhang: BUPT

Xiangyu Zeng: BUPT

Yang Yu: Central University of Fiance and Economics

Hui Tang: University of Science at Chongqing

'''Collaborative students'''

[http://lilt.cslt.org/ Lantian Li]

Askar Rouz: CSLT, VPR group

'''Collaborative researchers'''

Javier Tejedor, University of Madrid, Spain

Nick Evance, EURECOM, France

Simon King, University of Edinburgh, UK

Ravi Vipperla, Nuance, UK

Guozhen Zhao, Chinese Academy

'''Collaborative engineers'''

Xinkai Du, Huilan

Xiaoming Zheng, Sinovoice

'''Allumi'''

Jun Wang: CSLT, VPR group

Sheng Su: BUPT

Ke Ning: Beijing Institute of Technology

Haichao Yu: BUPT

Chao Yuan Zuo: Beijing Institute of Technology

Xi Ma: Beijing Post U.

Xizi Ouyang : Kuang yuan

Mahpirat: Xinjiang U.

Zulhumar, Xinjiang U.

Jun Qi, U. Washington

Hussein: Xinjiang U.

Fuhe Li: Petro Univ. Tsingdao

Tanche Li: CS, Tsinghua Univ.

Xuesheng Li: Petro Univ. Tsingdao

Xiangtao Meng: Chongqing Post U.

Xue Chen, Huilan

Zhao Fang, Huilan

Huabo Xie, Puqiang

Chao Liu

Bin Yuan, BUPT

Shi Yin: Chongqing Post U.

Yiye Lin: Beijing Institute of Technology

Fanhu Bie: CSLT, VPR group

Bo Hu, Puqiang

Mian Wang: BUPT

Personal Status

2015-12-16T13:50:23Z

Wangd：

[[Zhiyong Zhang]]

[[Mian Wang]]

[[Yiye Lin]]

[[Mengyuan Zhao]]

[[Xiangyu Zeng]]

[http://liuc.cslt.org/pages/status.html Chao Liu]

How to use SSH tunnel to access the campus network

2015-06-03T16:00:22Z

Wangd：

'''Goal:'''

Access the resources in the Tsinghua campus net, e.g., IEEE database.

'''Theory:'''

SSH provides a proxy function that enables connection of any TCP-based applications to be forwarded in or outside of a subnet. For more information:

http://www.yuexuan.org/?p=1000

'''How to: '''

* The only need is an account in a SSH server within the Tsinghua campus net. CSLT fellows please use 166.111.134.19.
* Following the instructions given by http://www.yuexuan.org/?p=1000, setup an SSH tunnel using putty with dynamic port, denoted by <port>, and then connect to our server
* If you are using cygwin or a linux client, simple ssh -D <port> 166.111.134.19
* Setup the browser to use the local machine (127.0.0.1) as the socket5 proxy. Set the port as the number <port> that you choose when setting the SSH tunnel.

How to use SSH tunnel to access the campus network

2015-06-03T15:58:58Z

Wangd：

'''Goal:'''

Access the resources in the Tsinghua campus net, e.g., IEEE database.

'''Theory:'''

SSH provides a proxy function that enables connection of any TCP-based applications to be forwarded in or outside of a subnet. For more information:

http://www.yuexuan.org/?p=1000

'''How to: '''

* The only need is an account in a SSH server within the Tsinghua campus net. CSLT fellows please use 166.111.134.19.
* Following the instructions given by http://www.yuexuan.org/?p=1000, setup an SSH tunnel using putty with dynamic port, and then connect to our server
* Setup the browser to use the local machine (127.0.0.1) as the socket5 proxy. Set the port as the number that you choose when setting the SSH tunnel.

How to use SSH tunnel to access the campus network

2015-06-03T15:56:41Z

Wangd：

'''Goal:'''

Access the resources in the Tsinghua campus net, e.g., IEEE database.

'''Theory:'''

SSH provides a proxy function that enables connection of any TCP-based applications to be forwarded in or outside of a subnet. For more information:

http://www.yuexuan.org/?p=1000

'''How to: '''

The only need is an account in a SSH server within the Tsinghua campus net.
CSLT fellows please use 166.111.134.19. Following the instructions given by http://www.yuexuan.org/?p=1000,
setup a tunnel using putty and connect to our server, and then setup the browser to use the local machine (127.0.0.1) as the socket5 proxy,
you can access the resources that are only available at Tsinghua.

How to access cvss from outside

2015-06-03T15:55:51Z

Wangd：

cvss can be access from the lab without any settings. If you want to connect from the outside, you need an SSH proxy.

1. set ssh proxy
With putty: select ssh->tunnel, click dynamic, and set source port as <your port>, click add, then connect to 166.111.134.198
With cygwin or linux client: ssh -D <your port> 166.111.134.19

2. set your browser
Set the socket 5 proxy in your browser as 127.0.0.1, port as <your port>.

3. connect to cvss as usual.

More details for ssh tunel:

http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/How_to_use_SSH_tunnel_to_access_the_campus_network

How to access cvss from outside

2015-05-25T15:22:24Z

Wangd：以“cvss can be access from the lab without any settings. If you want to connect from the outside, you need an SSH proxy. Setting the tunel by connecting to 166.111.134...”为内容创建页面

cvss can be access from the lab without any settings. If you want to connect from the outside, you need an SSH proxy.

Setting the tunel by connecting to 166.111.134.19 via SSH. The instruction is here:

http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/How_to_use_SSH_tunnel_to_access_the_campus_network

Then set the proxy in your browser, and access cvss.

Computing

2015-05-25T15:17:09Z

Wangd：

[[grid coumputing]]

[[steps of adding a new grid node]]

[[Use CSLT cluster]]

[[CSLT cluster nodes]]

[[CSLT Central Storage (CCS)]]

[[How to setup SGE]]

[[How to setup your homepage]]

[[How to use SSH tunnel to access the campus network]]

[[How to publish tools,data,code]]

[[How to access cvss from outside]]

Speech Technology Group

2015-04-24T02:00:32Z

Wangd：

== '''Welcome to Speech Recognition team in CSLT, RIIT, Tsinghua University.'''==

Research focus:

'''1. Speech recognition'''

We are interested in large vocubulary speech recognition, particularly on the modern topics such as noise robust ASR, deep belief network, covariance modeling etc.

'''2. Speech analysis'''

Our research focuses on speech signal analysis, such as sparse coding, non-negative matrix factorization, speech signal separation, noise cancelation etc.

'''3. Spoken document retrieval'''

We focus on key word spotting, spoken term detection, spoken document retrieval, FST-based indexing.

----

[[People]]

[[Porjects]]

[[Work progress]]

[[Time tables]]

----
[[ASR-news|News]]

[[Readings]]

----

[[asr-rules|Regulations]]

Test

2015-04-24T02:00:13Z

Wangd：以“xxxx”为内容创建页面

xxxx

Speech Technology Group

2015-04-24T02:00:01Z

Wangd：

Speech Technology Group

2015-04-24T01:59:47Z

Wangd：

Cslt-member-visitors

2015-04-23T09:02:19Z

Wangd：

==Professors==

===[[Li Yi (衣莉)]]===
[[文件:yili.png|150px]]

* Affiliation: U. of Agriculture
* 2014.9-
* Acoustics

===[[Jie Li (利节)]]===
[[文件:lijie.png|100px]]

* Affiliation: U. of Chongqing Science and Technology
* 2015.1-
* Complex network

==Engineers==

=== Rong Liu (刘荣) ===
* Affiliation: Huilan Info
* 2013.6-
* neural LM, dialog system, QA

==Students==

=== Chao Xing (邢超） ===
* Master
* Beijing Jiaotong University
* 2014.5 -
* text processing, distributed representation, machine translation
* Interested in machine learning and improve human intelligence by computer process.

=== Yiye Lin (林一叶）===
* Beijing Institute of Technology
* 2014.8-
* speech recognition, speaker recognition

===Xiangyu Zeng (曾翔宇)===
* Beijing Post. U.
* 2014.10-
* speech recognition

===Bin Yuan(袁彬)===
* Beijing Post. U.
* 2014.10-
* NLP

===Dongxu Zhang (张东旭)===
* Beijing Post. U.
* 2014.10-
* NLP

===Xi Ma (马习）===
* Beijing Post. & Tel. U.
* 2015.1-
* NLP

===Mian Wang (王冕）===
* Beijing Post. U.
* 2014.11-
* neural network visualization

===Zhiyuan Tang (汤志远）===
* CAS, Chengdu
* 2015.3-
* speech processing, neural computing

===Yiqiao Pan (潘一桥）===
* BUPT
* 2015.4-
* language processing

===Zhongwei Yao (幺忠伟）===
* BUPT
* 2015.4-
* speech processing, voice conversion

===Zipeng Ye (叶子鹏）===
* Tsinghua
* 2015.3-
* speech processing, speech synthesis

Cslt-emember-alumni

2015-04-23T08:59:16Z

Wangd：

==Faculties==
=== Yunqing Xia ===
[[文件:Yunqing sml.jpg|100px]]

'''Associate Research Professor [-2015.3]'''
'''Directory of Intelligent search Lab. '''
Ph.D: Institute of Computing Technology, Chinese Academy of Science 
 
'''Research Area''' 
NLP, computing linguistics, text mining 

== Students ==

=== Jiang Liu (刘建）===
* CSLT position: PhD student[2001-2013]
* Current position:

=== Lipeng Zhang (张利鹏）===
* CSLT position: Master student[2006-2009]
* Current position: TP-Link

=== Jiang Cao (曹犟）===
* CSLT position: Master student[2006-2009]
* Current position: Baidu

=== Junlan Pan (潘俊兰）===
* CSLT position: Master student[2006-2009]
* Current position:

=== Song Gao (高嵩）===
* CSLT position: Master student[2006-2009]
* Current position: Domintec
=== Gang Wang (王刚）===
* CSLT position: PhD student[2007-2011]
* Current position: d-Ear

=== Linlin Wang (王琳琳）===
* CSLT position: PhD student[2007-2013]
* Current position: Post doc. in Cambridge

=== Boyi Hao (郝博一）===
* CSLT position: Master student[2007-2010]
* Current position:

=== Gang Quan (全刚）===
* CSLT position: Master student[2007-2010]
* Current position: Baidu

=== Wenxiao Cao (曹文晓）===
* CSLT position: Master student[2007-2010]
* Current position:

=== Canhua Luo(罗灿华）===
* CSLT position: Master student[2008-2011]
* Current position: Post Savings Bank of China

=== Hongxian Wang(王宏显）===
* CSLT position: Master student[2008-2011]
* Current position: 163

=== Jue Hou(候珏）===
* CSLT position: Master student[2008-2011]
* Current position: Baidu

=== Liou Chen(陈丽欧）===
* CSLT position: Master student[2009-2012]
* Current position: Baidu

=== Chao Zhang (张超) ===
* CSLT position: Master student[2009-2012]
* Current position: PhD student in Cambridge

=== Chenhao Zhang (张陈昊） ===
* CSLT position: PhD student[2009-2014]
* Current position: ICBC

=== Junjun Wang (王俊俊） ===
* CSLT position: Master student[2010-2013]
* Current position: Bank of China

===[[member-tgy|Guoyu Tang (唐国瑜）]]===
* PhD student (2011.9-2015.3)
* NLP

=== Chen Gong （龚宸） ===
* CSLT position: Master student[2010-2013]
* Current position: China Aerospace Science & Industry Corp

=== Han Qiu (邱晗）===
* CSLT position: Master student[2011-2014]
* Current position: DeNA

==Collaborative Students==
=== Ouyang Xizi (欧阳夕子)===
* CSLT position: Visiting student[2012-2013]
* Current position:

=== Xiangtao Meng (孟祥涛)===
* CSLT position: Visiting student[2013-2014]
* Current position: Zheng zhou

===[[member-yinshi|Shi Yin (殷实)]]===
* Collaborative student (2014.4-2015.4)
* speech recognition, embedded system
* Current position: Huawei, Shenzhen

== Engineers ==

== Visitors ==

=== Xuewei Zhang (张雪薇) ===

* Shenyang Jianzhu University
* 2014.6-2014.10
* signal processing, robust speech recognition

=== Xi Ma (马习）===
* Beijing Post. & Tel. U.
* 2014.5-2014.9
* Language modeling

=== Le Liu (刘乐）===
* CSLT position: Visiting engineer [2013]
* Current position: engineer in d-ear

=== Fuhe Li (李富合）===
* CSLT position: Visiting student [2013.12-2014.5]
* Current position: master student in Beijing Post. U.

=== Xuesheng Li (李学生）===
* CSLT position: Visiting student [2013.12-2014.5]
* Current position:

=== Mahpirat (买哈铺热提•外力)===
* CSLT position: Visiting student [2013.1-2013.9]
* Current position:

=== Rayila Parhat (热依莱木•帕尔哈提) ===
* CSLT position: Visiting student [2013.1-2013.9]
* Current position:

=== Hexin Bai (白赫鑫) ===
* CSLT position: Visiting students [2013.12-2014.4]
* Current position:

=== Chen Xue (薛陈) ===
* CSLT position: Visiting engineer [2013.6-2014.7]
* Current position: Huilan Ltd.

===Hongzhi Yang (杨宏志) ===
* CSLT position: Visiting engineer [2014.3-2014.6]
* Current position: Puqiang Ltd.

=== Fang Zhao (赵芳) ===
* Affiliation: Huilan Ltd.
* 2013.9-2014.9
* QA, dialog system

Cslt-member-stu

2015-04-23T08:58:17Z

Wangd：

==CSLT Students==

===[[member-biefh|Fanhu Bie (别凡虎）]]===
* PhD student (2011.9-)
* speaker recognition

===[[member-wangjun|Jun Wang (王军）]]===
* PhD student (2011.9-)
* speaker recognition

===[[member-fanmiao|Miao Fan (范淼)]]===
* PhD student (2012.9-)
* NLP

===[[member-askar|Askar Rouz (爱斯卡尔.肉孜)]]===
* PhD student (2012.9-)
* speaker recognition

===[[member-lilantian|Lantian Li (李蓝天)]]===
* PhD student (2013.9-)
* speaker recognition

===[[member-liuchao|Chao Liu (刘超)]]===
* Master student (2012.6-)
* speech recognition

===[[member-xzd|Zhongda Xie (谢仲达)]]===
* Master student (2013.6-)
* NLP

===[[member-zxc|Xiaocong Zhou (周骁聪)]]===
* Master student (2014.9-)
* NLP

==Collaborative Students==

News-20150423

2015-04-23T08:36:34Z

Wangd：

CSLT released a free Uyghur speech database. This is the first free-downloaded speech database for Uyghur in the world. It can be used
to build a full-fledged Uyghur ASR system and speaker recognition system. Meanwhile, we publish two ASR challenges based on this database,
one is for clean speech and the other is for noisy data. We hope this will promote the speech research for Uyghur language.

Check the [http://cslt.riit.tsinghua.edu.cn:8081/data/thuyg20/README.html database homepage] for details.

News-20150423

2015-04-23T08:36:11Z

Wangd：

CSLT released a free Uyghur speech database. This is the first free-downloaded speech database for Uyghur in the world. It can be used
to build a full-fledged Uyghur ASR system and speaker recognition system. Meanwhile, we publish two ASR challenges based on this database,
one is for clean speech and the other is for noisy data. We hope this will promote the speech research for Uyghur language.

Check the database homepage for details:

[http://cslt.riit.tsinghua.edu.cn:8081/data/thuyg20/README.html click here]

News-20150423

2015-04-23T08:35:56Z

Wangd：以“CSLT released a free Uyghur speech database. This is the first free-downloaded speech database for Uyghur in the world. It can be used to build a full-fledged Uyghur...”为内容创建页面

CSLT released a free Uyghur speech database. This is the first free-downloaded speech database for Uyghur in the world. It can be used
to build a full-fledged Uyghur ASR system and speaker recognition system. Meanwhile, we publish two ASR challenges based on this database,
one is for clean speech and the other is for noisy data. We hope this will promote the speech research for Uyghur language.

Check the database homepage for details:

[http://cslt.riit.tsinghua.edu.cn:8081/data/thuyg20/README.html]

News

2015-04-23T08:31:17Z

Wangd：

[[news-20150423| Free Uyghur speech database THUGY-20 released]]

[[news-20150401| Prof. Nobuaki Minematsu from the University of Tokyo visit CSLT]]

[[news-20150126|Guorong Chen, Jie Li from CUST visit CSLT]]

[[news-20150104|VPR standard proposed by CSLT and collaborators published]]

[[news-20141119|Prof. Tanja Schultz from KIT Germany visit CSLT]]

[[news-20141021|Dr. Andrew Abel from University of Stirling, UK visit CSLT]]

[[news-20141015|Prof. Nick Evans from EURECOM visit CSLT]]

[[news-20140905|Conference attendance in Singapore]]

[[news-20140724|Prof. AlanBlack from CMU Visit CSLT]]

[[news-20140715|Prof. Woon-Seng Gan from NTU Visit CSLT]]

[[news-20140612|Dr. Frank Soong from Microsoft Asia Research Visit CSLT]]

[[news-20140526|IBM scientists visit CSLT]]

Publication-trp

2015-04-23T04:30:59Z

Wangd：

[[文件:vmfsne.png|200px]]
*[[媒体文件:Cslt-trp-template-vmfsne.pdf|TRP-20150009: VMF-SNE: Embedding for Spherical Data. Mian Wang, Dong Wang]]

[[文件:ros.png|200px]]
*[[媒体文件:Ros.pdf|TRP-20150008: Learning Speech Rate in Speech Recognition. Xiangyu Zeng, Shi Yin.]]

[[文件:Dnnvadstru.png|200px]]
*[[媒体文件:DNNVADTRP.pdf|TRP-20150007: Voice Activity Detection Based on Deep Neural Networks. Shi Yin.]]

[[文件:Uyghur-training.png|200px]]
*[[媒体文件:UyghurTRP.pdf|TRP-20150006: Low-resource Uyghur Acoustic Model Training based on Cross-lingual Features. Shi Yin.]]

[[文件:Beam-forming.png|200px]]
*[[媒体文件:Multi-Microphones_Reverberation_Cancellation_for_Distant_Speech_Recognition.pdf|TRP-20150005: Multi-Microphones_Reverberation_Cancellation_for_Distant_Speech_Recognition. Xuewei Zhang.]]

[[文件:Clipping-speaker.png|200px]]
*[[媒体文件:Clip.pdf|TRP-20150004: Detection and Reconstruction of Clipped Speech in Speaker Recognition. Fanhu Bie et al.]]

[[文件:Semi-dynamic-embedding.png|200px]]
*[[媒体文件:Taglm.pdf|TRP-20150003: Semi-Dynamic Graph Embedding for Large Scale Language Model Adaptation. Bin Yuan et al.]]

[[文件:Speaker-discriminative-score.png|200px]]
*[[媒体文件:DNN-based Discriminative Scoring for Speaker.pdf|TRP-20150002: DNN-based Discriminative Scoring for Speaker Recognition Based on i-vector. Jun Wang et al. ]]

[[文件:Noisy-traiing.png|200px]]
*[[媒体文件:Noisy Training for Deep Neural Networks in.pdf|TRP-20150001: Noisy Training for Deep Neural Networks in Speech Recognition. Shi Yin et al.]]

[[文件:English-scroing.png|200px]]
*[[媒体文件:AutomaticScoringforEnglishUtterances.pdf|TRP-20140001: Automatic Scoring for English Utterances. Bo Hu.]]

[[媒体文件:Template.rar|CSLT-TRP latex template]]

文件:Vmfsne.png

2015-04-23T04:30:47Z

Wangd：

文件:Cslt-trp-template-vmfsne.pdf

2015-04-23T04:28:31Z

Wangd：

People you may want to follow

2015-04-21T01:25:55Z

Wangd：

1. Dong Yu, MS

http://research.microsoft.com/en-us/people/dongyu/

2. Geoffrey E. Hinton, U. Toronto

http://www.cs.toronto.edu/~hinton/

3. Deng Li, MS

http://research.microsoft.com/en-us/people/deng/

4. Tara Sainath, IBM

http://www.researchgate.net/profile/Tara_Sainath/publications/

http://www.informatik.uni-trier.de/~ley/pers/hd/s/Sainath:Tara_N=.html

5. Daniel Povey, JHU

http://www.danielpovey.com/

6. Patrick Kenny, CRIM

http://www.crim.ca/perso/patrick.kenny/

7. Nelson Morgan, ICSI Berkley

http://www1.icsi.berkeley.edu/~morgan/

8. Jonathan Berant , Stanford NLP

http://www-nlp.stanford.edu/joberant/

9. Yann LeCun

http://yann.lecun.com/

10. Hua Wang

http://inside.mines.edu/~huawang/

11. Heng Huang

http://ranger.uta.edu/~heng/publications.html

12. Feipeng Nie

http://www.escience.cn/people/fpnie/papers.html

13. Yoshua Bengio

http://www.iro.umontreal.ca/~bengioy/yoshua_en/

14. Nils J. Nilsson

http://ai.stanford.edu/~nilsson/

15. James Martens

http://www.cs.toronto.edu/~jmartens/research.html

16. Hui Jiang

https://wiki.eecs.yorku.ca/user/hj/

Other teams

2015-03-29T14:58:49Z

Wangd：

[http://nlp.stanford.edu/read/ Standford NLP team reading group]

[http://deeplearning.net/ Deep learning group]

Other teams

2015-03-01T12:33:12Z

Wangd：以“[http://nlp.stanford.edu/read/ Standford NLP team reading group]”为内容创建页面

[http://nlp.stanford.edu/read/ Standford NLP team reading group]

Readings

2015-03-01T12:32:47Z

Wangd：

[[New member reading list]]

[[Advanced member reading list]]

[[Public Research Tools]]

[[People you may want to follow]]

[[Other teams]]

[[Journals you may submit to]]

[[Free libraries]]

[[Data resources]]

Visiting-events-2015-01-24

2015-02-03T10:55:45Z

Wangd：

==Plan==

Plan:

2015-01-30, driving at 10:00am, distribute presents, donate, return to Beijing at 2:pm

Present list:

children toys, food, daily usage, etc.

==Event==

===Information===

Time: 2015-1-30

Attendees: 衣莉（农大）韩瑜（大都会理财）赵梦原（清华）

===donation===
{| class="wikitable"
!doner !!donation !! 花费 !! 其他
|-
|毕蕾蕾 ||一大包玩具 ||2166 ||详见账单
|-
|冯兰、王兰芳 ||每次帮助签收货物，30号帮助搬运货物 ||4334 ||捐款
|-
|韩瑜|| 一大包玩具，来回4个小时车程去廊坊晋元道26号牧羊地儿童村。看望孤儿和搬运货物
|-
|霍东烨|| 30瓶洗发水，3大瓶沐浴乳，10袋小米
|-
|李颖异 ||玩具、30号帮助搬运货物
|-
|李蔚、王柯|| 200元
|-
|Max ||1000元
|-
|清华语音研究中心 ||1000元
|-
|沈曦 ||1000元
|-
|唐莹 ||1800元
|-
|衣莉 ||1500元
|-
|张洁 ||一大包玩具
|-
|赵梦原 ||30号搬运主要劳动力，一起去看望小朋友们
|-
|}

===Purchase list===

[[媒体文件:购物清单-1.pdf|购物清单]]

===Receipt===

[[文件:捐款收据.jpg|捐款收据|500px]]

Pictures:

[[文件:外景.jpg|外景|800px]]
[[文件:门牌.jpg|门牌|800px]]
[[文件:照片墙1.jpg|照片墙1|800px]]
[[文件:照片墙2.jpg|照片墙2|800px]]
[[文件:物资.jpg|物资|800px]]
[[文件:资助企业.jpg|资助企业|800px]]
[[文件:资助个人.jpg|资助个人|800px]]
[[文件:张强.jpg|张强|800px]]
[[文件:衣老师和孩子.jpg|衣老师和孩子|800px]]
[[文件:小学.jpg|小学|800px]]
[[文件:画.jpg|画|800px]]
[[文件:教室1.jpg|教室1|800px]]
[[文件:作业.jpg|作业|800px]]
[[文件:图书室.jpg|图书室|800px]]
[[文件:画室.jpg|画室|800px]]

Feedback:

ASR-news-201501012

2015-02-03T10:49:20Z

Wangd：

孟祥涛和张茜2015年2月1日大婚，祝贺语音组第一件新婚之喜！小孟为之勇、晓曦等做出了榜样！

[[文件:20150201 115728.jpg|表白|500px]]

[[文件:IMG 1209.JPG|合影|500px]]

How to use SSH tunnel to access the campus network

2015-01-27T00:26:36Z

Wangd：以“'''Goal:''' Access the resources in the Tsinghua campus net, e.g., IEEE database. '''Theory:''' SSH provides a proxy function that enables connection of any TCP-...”为内容创建页面

'''Goal:'''

Access the resources in the Tsinghua campus net, e.g., IEEE database.

'''Theory:'''

SSH provides a proxy function that enables connection of any TCP-based applications to be forwarded in or outside of a subnet. For more information:

http://www.yuexuan.org/?p=1000

'''How to: '''

The only need is an account in a SSH server within the Tsinghua campus net. CSLT fellows please use 166.111.134.19. Following the instructions given by http://www.yuexuan.org/?p=1000, setup a tunnel using putty and connect to our server, and then setup the browser to use the local machine (127.0.0.1) as the socket5 proxy, you can access the resources that are only available at Tsinghua.