Home > People > Zhanjiang Song

Zhanjiang Song
Member of Research Staff
NRC Beijing

 

I joined Nokia Research Center in Beijing as Senior Research Engineer in May, 2007. Since then, I have been working on service-oriented language processing technologies and applications, as well as end-to-end service architectures for scalable rich context data processing systems.

Prior to joining Nokia, I had worked for about 1 year as research scientist in a company in Hong Kong, and more than 5 years as co-founder and R&D VP in a start-up company in Beijing, both focusing on speech and language processing systems.

I received my B.S. degree in 1994, majoring in Computer Software and M.S. degree in 1997, majoring in Computer Network, both from Nankai University, China. I received my Ph.D degree in 2001, majoring in Speech Recognition and Language Understanding, from Tsinghua University, China.

 

Professional Activities

  • From Jul. to Aug., 2000, I participate in the international research project (in series) "JHU Workshop 2000: Pronunciation Modeling of Mandarin Casual Speech" carried out at Johns Hopkins University, USA (http://www.clsp.jhu.edu/ws2000/groups/mcs)
  • I was a member of "Chinese Speech Interactive Technology Standard Group (CSITSG)" established by the Ministry of Information Industry (MII) of China, and participated in drafting of "Technical Standard for Automatic Voiceprint Recognition (Speaker Recognition)" (http://www.speechstandard.org.cn)
 

Research Interests

My research interests include: 

  • Speech and language technologies, such as speech recognition, speaker recognition, natural language processing, etc.
  • Software engineering, such as architecture and interface design and implementation, system integration, etc.
  • Computer network and distributed systems, such as distributed computing and data stream mangement system, etc.
 

Publications

Selected publications in Chinese:

  • Zhang J-Y, Zheng F, Du S, Song Z-J, Xu M-X. Merging-Based Syllable Detection Automaton in Continuous Speech Recognition. Journal of Software. 10(11): 1212~ 1215. Nov. 1999 (in Chinese)
  • Song Z-J, Zheng F, Xu M-X, Wu J, Wu W-H. Research on Chinese Continuous Speech Recognition System and Knowledge Based Search Strategies. Acta Automatica Sinica. 26(4): 470~477. Jul. 2000 (in Chinese)
  • Xiong Z-Y, Zheng F, Song Z-J, Wu W-H. Tree-Structural Universal Background Model Based Efficient Speaker Identification. Journal of Tsinghua University (Science & Technology). 46(7): 1305~1308. Jul. 2006 (in Chinese)

Selected publications in English:

  • Song Z-J, Xu J-D, Wu G-Y. A Distributed Implementation Method on Reliable Network Broadcast Communications Environment. In Proc. 1st International Conference on Information Infrastructure (ICII'96). Beijing, China, Apr., 1996
  • Zheng F, Song Z-J, Li L et al. The Distance Measure for Line Spectrum Pairs Applied to Speech Recognition. In Proc. International Conference on Spoken Language Processing (ICSLP). 1998, 3: 1123~1126
  • Song Z-J, Zheng F, Xu M-X, Wu W-H. An Effective Scoring Method for Speaking Skill Evaluation System. In Proc. European Conference on Speech Communication and Technology (EuroSpeech). 1999, 1: 187~190
  • Zheng F, Song Z-J, Xu M-X et al. EasyTalk: a Large-Vocabulary Speaker-Independent Chinese Dictation Machine. In Proc. European Conference on Speech Communication and Technology (EuroSpeech). 1999, 2: 819~822
  • Song Z-J, Zheng F, Wu W-H. Statistical Knowledge Based Frame Synchronous Search Strategies in Continuous Speech Recognition. In: International Conference on Acoustics, Speech and Signal Processing (ICASSP). 2000, 3: 1583~1586
  • Zheng F, Wu J, Song Z-J. Improving the Syllable-Synchronous Network Search Algorithm for Word Decoding in Continuous Chinese Speech Recognition. Journal of Computer Science and Technology (JCST). 2000, 15(5): 461~471
  • Fung P, Byrne W, Zheng F, Kamm T, Liu Y, Song Z-J, Venkataramani V, Ruhi U. Pronunciation Modeling of Mandarin Casual Speech. Final Report for Workshop 2000 for Language Engineering for Students and Professionals Integrating Research and Education. http://www.clsp.jhu.edu/ws2000/final_reports/mpm/.
  • Zheng F, Song Z-J, Fung P, Byrne W. Reducing Pronunciation Lexicon Confusion and Using More Data without Phonetic Transcription for Pronunciation Modeling. In Proc. Int. Conf. on Spoken Language Processing (ICSLP). 2002. 2461~2464. Sep. 16-20, 2002, Colorado, USA
  • Xiong Z-Y, Zheng F, Song Z-J, Wu W-H. Combining Selection Tree with Observation Reordering Pruning for Efficient Speaker Identification Using GMM-UBM. In Proc. Int. Conference on Acoustics, Speech and Signal Processing (ICASSP). I: 625~628. Mar. 19-23, 2005. Philadelphia, USA
  • Xiong Z-Y, Zheng F, Song Z-J, Soong F, Wu W-H. A Tree-Based Kernel Selection Approach to Efficient Gaussian Mixture Model - Universal Background Model Based Speaker Identification. Speech Communication. 48(10): 1273~1282. Oct. 2006
 

Patents

4 granted patents, 1 filed invention report.