电子工程系

Department of Electronic Engineering

    吴及 ,博士 副教授

    中国北京市清华大学电子工程系 100084
    电话: +86-10-62781706
    传真:+86-10-62770317
    电子邮箱: wuji_ee@tsinghua.edu.cn


 

 

    吴及,清华大学电子工程系,副教授,博士生导师。

    1991年进入清华大学学习,分别于1996年和2001年在清华大学电子工程系获得学士和工学博士学位,后留校任教。2005年晋升副教授,2009年获得博士生导师资格。2013年9月至2015年8月在美国佐治亚理工学院担任访问学者。主要从事数据结构与算法方面的教学工作,以及智能语音和语言技术,机器学习,大数据等领域的研究工作。承担以及作为骨干成员参加了863,国家自然科学基金,工信部电子发展基金等多项国家科研项目。从2006担任清华-讯飞语音技术联合实验室主任。目前是中国语音产业联盟技术工作组组长。参加的项目“智能语音交互关键技术及应用开发平台”于2011年获国家科技进步二等奖。负责的项目“面向海量语音数据的识别、检索和内容分析技术及其应用”获2014年度北京市科学技术奖一等奖。已在国内外刊物和学术会议上发表论文九十余篇,包括IEEE Trans on ASLP,IEEE Trans. on SMC,IEEE Signal Processing Letters,Neural Network World,清华大学学报,电子与信息学报, 模式识别与人工智能,ICASSP,InterSpeech等重要学术期刊和学术会议。2009年起为全国人机语音通讯学术会议常设机构委员,担任国内外多个期刊和学术会议的审稿人,现在为IEEE高级会员。

 

教育背景

1996年毕业于清华大学电子工程系无线电技术与信息系统专业,获工学学士学位

2001年于清华大学电子工程系获得信号与信息处理专业博士学位

 

工作履历

2001至今 清华大学电子工程系

学术兼职

2009年8月-至今  全国人机语音通讯学术会议常设机构委员会委员(NCMMSC Standing Committee)

ISCSLP2012,ISCSLP2010,ISCSLP2008程序委员会委员

NCMMSC2013,NCMMSC2011,NCMMSC2009,NCMMSC2007程序委员会委员

 

社会兼职

2012年8月至今  中国语音产业联盟  技术工作组组长

2004年4月至今  工业与信息化部中文语音交互技术标准工作组   成员

 

研究领域

语音识别和人机交互

基于内容的语音分析和检索

自然语言处理

数据挖掘,机器学习和模式识别

 

研究概况

2016.1-2019.12:国家自然科学基金面上项目“音频事件检测技术研究”,61571266,项目负责人

2012.1-2015.12:国家自然科学基金面上项目“中文自动口语摘要技术研究”,61170197, 项目负责人

2012.9-2014.12.30:863计划项目子课题“海量非结构化数据的集成管理和分析,舆情分析示范应用”,2012AA011004,研究骨干

2012.6-2015.6:清华-讯飞语音技术联合实验室(三期),安徽科大讯飞信息科技股份有限公司

2009.5-2012.5:清华-讯飞语音技术联合实验室(二期),安徽科大讯飞信息科技股份有限公司

2006.2-2009.2:清华-讯飞语音技术联合实验室(一期),安徽科大讯飞信息科技股份有限公司 

2012.2- 2013.2:语音识别联合研发项目,腾讯科技(深圳)有限公司 

2012.8- 2012.12:文本库采集及语言模型训练,北京三星通信技术研究有限公司 

2006.6.1-2010.10.31:863计划十一五重点项目“多语言语音合成关键技术研究与应用产品开发“子课题“基于统计建模的个性化语音合成技术研究”,2006AA010104,课题负责人

2006.11-2008.12:“863”面上项目探索导向类课题“基于内容的高性能语音搜索技术探索研究”,2006AA01Z149,项目负责人,

2004.4-2005.5:鲁棒语音识别技术研究,北京东芝研究中心

2001-2003:“863”计划项目“智能化中文语音信息处理平台”,2001AA114071,项目负责人

 

奖励与荣誉

2014  面向海量语音数据的识别、检索和内容分析技术研发及应用,北京市科学技术奖,一等奖

2011  国家科学技术进步二等奖(个人排名:第8)

学术成果

Journal Papers:

[1] Wu J, Li M, Lee C H. A Probabilistic Framework for Representing Dialog Systems and Entropy-Based Dialog Management through Dynamic Stochastic State Evolution[J]. Audio, Speech, and Language Processing, IEEE Transactions on, 2015, 23(11): 2026 – 2035
[2] Zhiyang He, Ji Wu,Tao Li. Label Correlation Mixture Model: A Supervised Generative Approach to Multilabel Spoken Document Categorization[J]. Emerging Topics in Computing, IEEE Transactions on, 2015, 3(2): 697-710
[3] Xiao-Lei Zhang and Ji Wu, “Deep belief networks based voice activity detection,” IEEE Transactions on Audio, Speech, and Language Processing, vol. 21, no. 4, 697-710, April 2013.
[4] X.-L. Zhang and J. Wu, “Linearithmic time sparse and convex maximum margin clustering,” IEEE Transactions on Systems, Man, and Cybernetics—Part B: Cybernetics, vol. 42, no.6, pp. 1669-1692, 2012.
[5] Wu Ji, Zhang Xiao-Lei. SPARSE KERNEL MAXIMUM MARGIN CLUSTERING. Source: NEURAL NETWORK WORLD, v 21, n 6, p 551-573.2011. (IDS Number: 879TT,EI: 20120614740409)
[6] Wu Ji, Zhang Xiaolei. Efficient multiple kernel support vector machine based voice activity detection, In: Proc. IEEE Signal Process Letters, 2011,18(8), pp466-469. (IDS Number: 783QM,EI: 20112714110530)
[7] Wu Ji, Zhang Xiaolei. An efficient voice activity detection algorithm by combining statistical model and energy detection, EURASIP Journal on Advances in Signal Processing, 2011,2011(1), pp18-27.(IDS Number: 819TL)
[8] Wu Ji, Zhang Xiaolei. Maximum margin clustering based statistical VAD with multiple observation compound feature, IEEE Signal Process. Lett., 2011,18(5), pp283-286. (IDS Number: 734LC, EI: 20111113757684)
[9] Li Wei, Wu Ji, Lv Ping. Query Expansion Based High Performance Chinese Voice Retrieval, Pattern Recognition and Artificial Intelligence, 2011,8, 24,(4),pp561-566. Language: Chinese. EI: 20114114415342
[10] Zhang Xiaolei, Wu Ji, Lv Ping. Support vector machine based VAD by using multiple observation compound feature, Journal of Tsinghua University(Science and Technology) 2011, 51(9),pp1209-1214. Language: Chinese. EI: 20114514493905
[11] Su Tengrong, Wu Ji, Wang Zuoying. Acoustic Model Training Based on Spatial Correlation Transformation, Journal of Electronics & Information Technology, Journal of Electronics & Information Technology. 2010, 32(4), 1003-1007. Language: Chinese. EI: 20102112949946
[12] Su Tengrong, Wu Ji, Wang Zuoying. Spatial correlation transformation for speech recognition, Journal of Tsinghua University, v 49, n 10, p 1655-1659, October 2009. Language: Chinese. EI: 20094612457047
[13] Li Wei, Wu Ji, Wang Zhiguo. Fast lattice generation algorithm, Journal of Tsinghua University(Science and Technology). 2009, 49(S1), pp1254-1257. Language: Chinese.
[14] Qian Sheng, Lv Ping, Wu Ji. Maximum probability increase estimation method for fast Gaussian likelihood computations. Journal of Tsinghua University(Science and Technology). 2009, 49(S1), 1258-1261. Language: Chinese.
[15] Su Tengrong, Wu Ji, Wang Zuoying. Spatial correlation transformation for speech recognition. Journal of Tsinghua University(Science and Technology). 2009, 49(10), pp82-86. Language: Chinese.
[16] Wang Zhiguo, Wu Ji, Wang Renhua, Dai Lirong. An Algorithm of Model Compensation Based on The Estimation of Additive Noise and Channel Function for Speech Recognition. ACTA ACUSTICA,2008,33(3), pp 238-243. Language: Chinese. (EI: 20082311304376)
[17] Hu Yanfang, Wu Ji, Liu Huixing. Speech/music Discrimination Based on A Modified Low Energy Ratio. Journal of Tsinghua University(Science and Technology). 2008,48(1), pp720-724. Language: Chinese. (EI: 20082811370592)
[18] Wu Ji, Xiao Xi, Xu Lin, Wang Zuoying. Hidden Markov Model Search Algorithm for Non-convex Duration distributions. Journal of Tsinghua University(Science and Technology). 2005,45(7), pp.924-927. Language: Chinese. (EI: 2005369347292)
[19] Wu Fengliang, Wu Ji, Wang Zuoying. Exponential Threshold Based Speech Endpoint Detection Method. Journal of Data Acquisition & Processing. 2005, 20(4), pp.385-389. Language: Chinese. (EI: 2006049664384)
[20] Chen Junyan, Wu Ji, Wang Xia. Robust Language Understanding Algorithm in Spoken Dialogue Systems. Journal of Tsinghua University(Science and Technology). 2005,45(1), pp21-24. Language: Chinese. (EI: 2005139014221)
[21] Ji Wu, Zuoying Wang. An Efficient Computation Algorithm in Mandarin Continuous Speech Recognition. Chinese Journal of Electronics, vol11, n1, 2002, pp44-47, (IDS Number: 516QH, EI: 2002156912926)
[22] Lv Ping, Wu Ji, Wang Zuoying, Lu Dajin. Rapid speaker adaptation for continuous speech recognition. Journal of Tsinghua University(Science and Technology). 2002, 42(7),pp.977-980. Language: Chinese. (EI: 2002427145372)
[23] Ji Wu, Zuoying Wang. A Decision Tree-Structured Algorithm of Speaker Adaptation Based on Gaussian Similarity Analysis. Chinese Journal of Electronics, 2001, 10(2): 166-169. (IDS Number: 428WU, EI: 2001256553266)
[24] Zuoying Wang, Ji Wu. A New Similarity Measure of Random Variables and Its Application to VQ for Speech Recognition. Chinese Journal of Electronics, 2000, 9(4),pp.448-452[4]. (IDS Number: 384TJ. EI: 2001015411506.)
[25] Wu Ji, Liu Feng, Wang Zuoying. Research on optimal algorithm of likelihood computation in continuous speech recognition. Journal of Tsinghua University(Science and Technology). 1999,39(5),pp.77-80.(EI: 2000215118517). Language: Chinese.

Conference Papers:

[1] Wu J, Li M, Lee C H. An Entropy Minimization Framework for Goal-Driven Dialogue Management[C]//Sixteenth Annual Conference of the International Speech Communication Association. 2015
[2] Zhen Huang, Jinyu Li, Sabato Marco Siniscalchi, I-Fan Chen, Ji Wu and Chin-Hui Lee.Rapid Adaptation for Deep Neural Networks through Multi-Task Learning. Sixteenth Annual Conference of the International Speech Communication Association. 2015
[3] Hongyi Ding, Ji Wu. Predicting Retweet Scale Using Log-Normal Distribution. BigMM 2015. Beijing, 2015. 4
[4] Zhiyang He, Ji Wu, Ping Lv.Label Correlation Mixture Model for Multi-label Text Categorization. SLT 2014,South Lake Tahoe,2014.12, pp.83-88
[5] Zhipeng Chen, Teng Zhang, Ji Wu. Subword Scheme for Keyword Search. IEEE Workshop on SLT 2014, South Lake Tahoe, USA, 2014.12, pp.483-488
[6] Teng Zhang, Ji Wu, Dingding Wang,Tao Li,"Audio Retrieval Based on Perceptual Similarity.”collaborateCom,2014.
[7] Zhiyang He, Ping Lv, Ji Wu.An Effective and Robust Approach to Mandarin Spoken Language Understanding in Specific Domain. ISCSLP 2014,Singapore,2014.9, pp.604-608.
[8] Zhiyang He, Ping Lv, Ji Wu.Minimum Classification Error Rate Training of Supervised Topic Mixture Model for Multi-label Text Categorization. ISCSLP 2014,Singapore,2014.9, pp.39-43.
[9] Zhipeng Chen, Zhiyang He, Ping Lv, Ji Wu.Improving Keyword Search by Query Expansion in a Probabilistic Framework. ISCSLP 2014, Singapore, 2014.9, pp.187-191.
[10] Miao Li, Hongyi Ding, Ji Wu,Global Discriminative Model for Dependency Parsing in NLP Pipeline. ISCSLP 2014,Singapore,2014.9, pp.614-618.
[11] Shusen Li, Zhiyang He, Ji Wu.An Ontology Semantic Tree based Natural Language Interface. ISCSLP 2014,Singapore,2014.9, pp.226-230.
[12] X.-L. Zhang and J. Wu. "Denoising deep neural networks based voice activity detection," In Proceedings of the 38th IEEE International Conference on Acoustic, Speech, and Signal Processing, Vancouver, Canada, May, 2013, pp. 853-857.
[13] X.-L. Zhang and J. Wu. "Weight optimization and layered clustering-based ECOC," In Proceedings of the 38th IEEE International Conference on Acoustic, Speech, and Signal Processing, Vancouver, Canada, May, 2013, pp. 3477-3481.
[14] Zhiyang He, Ping Lv, Wei Li, Wu Ji, A Synchronized Pruning Composition Algorithm of Weighted Finite State Transducers for Large Vocabulary Speech Recognition. the 8th International Symposium on Chinese Spoken Language Processing (ISCSLP2012) , Hongkong, 2012.12. pp11-15.
[15] Qinghua Wu, Zhang Xiao-Lei, Ping Lv, Wu Ji, Perceptual Similarity between Audio Clips and Feature Selection for Its Measurement. the 8th International Symposium on Chinese Spoken Language Processing (ISCSLP2012) , Hongkong, 2012.12. pp387-391.
[16] X. L. Zhang, J. Wu, Z. P. Chen, and P. Lv, “Optimized weighted decoding for error correcting output codes,” in Proceedings of the  International Conference on Acoustic, Speech, and Signal Processing (ICASSP), 2012, pp. 2101–2104.
[17] Zhihui Du, Xiangyu Li, Ji Wu. Accelerating the Training of HTK on GPU with CUDA. ParLearning 2012. Shanghai, China, May, 2012
[18] Wu Ji, He Zhiyang, Lv Ping. An Active Learning Approach to Task Adaptation,in Proc. Interspeech 2011, Florence, Italy, 2011, pp2597-2600.
[19] Wei Li, Ji Wu, Ping Lv. High Performance Chinese Spoken Term Detection Based on Term Expansion. ISCSLP 2010, Tainan, TaiWan, 2010, pp. 430-434, EI: 20110713663225
[20] Wenzhu SHEN, Ji WU and Wei LI. Web-Based Keyword Adapted Language Modeling for Keyword Spotting ISCSLP 2010, Tainan, TaiWan, 2010, pp. 251-255, EI: 20110713663270
[21] J. Wu, X. Zhang, and W. Li, “A New VAD Framework Using Statistical Model and Human Knowledge Based Empirical Rule,” in Proceedings of the 11th annual conference of the ISCA (INTERSPEECH 2010), Sept. 2010, pp. 3090–3093, EI: 20112714117442
[22] Wenzhu Shen, Roger Peng Yu, Frank Seide, Ji Wu. Automatic Punctuation Generation For Speech, In: Proc. IEEE Automatic Speech Recognition and Understanding Workshop(ASRU), Merano, Italy, 2009.12, p 586-589,EI: 20101112773312
[23] Tengrong Su, Ji Wu, Zuoying Wang, Jie Hao. Improvements on minimum covariance based spatial correlation transformation, In: ICASSP, Taipei, 2009.4, p4581-4584, EI: 20093912339320
[24] Tengrong Su,Ji Wu, Zuoying Wang. Spatial Correlation Transformation Based on Minimum Covariance, In: Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Las Vegas, USA, 2008,4. (EI: 20083811562338, ISTP : BHY47)
[25] Wei Li,Ji Wu,Zhiguo Wang. A Trellis Based Fast Lattice Generating Algorithm. In Proc. International Symposium on Chinese Spoken Language Processing(ISCSLP), Kunming, 2008,12, pp.189-192. (EI: 20091011939065,ISTP: BJA83)
[26] Zhiyang He,Zhiguo Wang,Wei Li,Ji Wu. A Combined Task Analysis Method for Data Selection in Mandarin Isolated Word Recognition System. In Proc. International Symposium on Chinese Spoken Language Processing(ISCSLP), Kunming, 2008,12, pp.213-216.(EI: 20091011940617, ISTP: BJA83)
[27] Zhiyang He,Wei Li,Ji Wu. Task analysis methods for data selection in task adaptation on Mandarin isolated word recognition. In: Proc. International Conference on Signal Processing (ICSP), Beijing, P.R.C, 2008,10. pp. 697-700. EI: 20092612148926
[28] Chi Zhang,Ji Wu,Xi Xiao, Zuoying Wang. Pronunciation Variation Modeling for Mandarin with Accent, In Proc. of Interspeech-ICSLP2006,Pittsburgh, Sep. 2006,pp.709-712. Language: Chinese. (EI: 20082511318873)
[29] Yizhou Wang, Ji Wu, Zuoying Wang. A two-dimensional robust algorithm based on Mel-bank Log-spectrum. In Proc of International Symposium on Communications and Information Technologies 2005, Beijing, pp.734-738.(EI: 20064310195136, ISTP: BDY20)
[30] Ye Tian, Ji Wu, Zuoying Wang, Dajin Lu. Fuzzy Clustering And Bayesian Information Criterion Based Threshold Estimation For Robust Voice Activity Detection. In: Proc of International Conference on Acoustics, Speech and Signal Processing(ICASSP). 2003. (EI: 2003397648772)
[31] Junyan Chen, Ji Wu, Zuoying Wang. A Chinese spoken dialogue system for train information. In: Proc of IEEE SMC’2003 [C]. Washington D.C., USA, 2003 (EI: 2003487750883)
[32] Ye Tian, Ji Wu, Zuoying Wang, Dajin Lu. Robust Noisy Speech Recognition with Adaptive Frequency Bank Selection. In: Proc of IEEE fourth International Conference on Multimodal Interfaces (ICMI'2002), Pittsburgh, 2002. (ISTP: BV52J)
[33] Ji Wu, Zuoying Wang. Gassian Similarity Analysis and Its Application in Speaker Adaptation. In: Proceeding of ICSLP2000, IV. pp.370-373.