Guojun Zhou
14670 NW Cornell Rd., Apt 48 Tel:
(503) 264-1700
Portland, OR 97229 Email: gzhou_us@yahoo.com
SUMMARY
- Expertise in speech
recognition, text-to-speech (TTS), spoken dialog design, and Digital
Signal Processing (DSP)
- Hands-on experience of
DSP algorithm optimization for microprocessor Architecture
- Experienced with
software engineering methodology
- Strong computer
programming and debugging skills (C/C++, Windows, Unix, etc.)
- Significant strengths
in project planning & execution, and communication skills
- Responsible for eight
U.S. and international patents
- Three years’ experience
with board-level circuit design and test
·
Highly reliable, sensible, well-organized contributor in team environment
INDUSTRY EXPERIENCE
Sr.
Software Engineer & Staff System Performance
Architect
Intel Corporation, Hillsboro, OR (1999 to present)
- Major researcher of new
DSP algorithms: improved the
robustness of speech recognition. Results published at primary
international conference. (see Publications List)
- Chief optimization
architect: built Intel Integrated Performance Primitive (IPP) Library of
DSP algorithms for Intel handheld microprocessor architecture. Provided
framework for team use of optimization and benchmarking.
·
Server performance architect: project performance for
Intel future server architectures to provide guidance for architects and
performance data for marketing and other executives.
- Major designer and
developer: built two Intel spoken dialog systems. Demonstrations helped
Intel gain publicity and influence the next generation human-machine
interfaces.
- Invited by other Intel
units to provide analysis of speech recognition algorithm for Intel
streaming processor architecture. Results impacted the architecture’s
roadmap.
- Invited speaker at
international conferences. Promoted Intel products and research to
academic and industry world to facilitate industry efficiency and
influence standards.
- Sole technical liaison
between Intel and two major U.S. universities (MIT and University of
Colorado at Boulder). Acquired the most advanced research results for
Intel.
Software
Engineer (Intern) Nuance Communications, Menlo
Park, CA (1998)
- Implemented
decision-tree based clustering algorithm in C++, which resulted in a more
reliable Nuance 6.0 speech recognition engine, a pioneer in telephony
area.
Senior
Hardware Engineer Thakral Technology Pte Ltd,
Singapore (1996)
- Ensured VCD player
product to reach market on time by leading efforts between design team in
Singapore and manufacturing group in China.
- Significantly increased
the sale of Panasonic and Sharp Video Compact Disk (VCD) players in China
by modifying their hardware circuits.
Project
Engineer Gateway
Technologies Services Pte Ltd, Singapore (1994 to 1996)
- Initiated speech
research and established the R&D environment for speech recognition.
·
Completed project design and development in Singapore English
recognition. Enabled the team to do further research and development work in
speech processing.
Software Engineer Xinhe Electronics Inc., Beijing, China (1993
to 1994)
- Designed Chinese speech recognition algorithms. The new algorithms
improved the recognition accuracy and enabled the technology to be used in
telephony area.
Hardware Engineer Xinhe Electronic Audio Co., Guangdong, China (1988 to 1990)
- Designed electronic circuits for a model of Xinhe audio systems,
which was awarded the Gold Medal at the 1991 Beijing International
Consumer Products Fair.
- Ensured product quality and availability in market on time by
leading standard compliance testing and efforts between design team and
manufacturing facilities.
PATENT
- R&D work at Intel
from 2000-2001 resulted in eight U.S. and international patents.
ACADEMIC
EXPERIENCE
Adjunct
Instructor Portland State University, Portland, OR (2001 to present)
- Teach a graduate course
titled: “Human-Computer Voice Communications”.
Professional
Research Assistant University of Colorado,
Boulder, CO (1999)
- Formulated new
adaptation techniques for speech recognition using DSP algorithms.
- Developed novel
acoustical features for speaker verification.
Research
Assistant Duke University, Durham, NC (1996 to 1998)
·
Conducted
research on speech signal processing and applied results to automatic speech
recognition and stress classification for aircraft cockpit recordings. (see
Publications List)
·
Developed a telephony-based data collection station for speaker
verification system.
·
Investigated new speech enhancement algorithm using auditory masking concepts.
Research
Assistant Tsinghua University, Beijing, China (1990 to 1993)
- Designed a speech
recognition system using Hidden Markov Model (HMM) algorithm. (see
Publications List)
- Completed a
voice-automated PBX (Private Branch Exchange) operator system.
- Designed hardware
interfaces between PC and PBX systems.
Research
Assistant Southeast
University, Nanjing, China (1987 to
1988)
·
Designed hardware circuits based on DSP algorithms for a digital
adaptive filter used for canceling interference in spread spectrum
communication systems.
EDUCATION
- Ph.D. Electrical
Engineering Duke University,
Durham, North Carolina, U.S.A.
- M.S. Electrical
Engineering Tsinghua University, Beijing, China
- B.S. Electrical
Engineering Southeast University, Nanjing, China
COMPUTER & OTHER SKILLS
- Operating Systems:
MS-Windows UNIX, LINUX, MS-DOS,
- Programming Languages:
C/C++, Perl, C-Script, x86/ARM Assembly, Java, HTML
- Software: Matlab,
AutoCAD, ESPS, xwaves, HTK, Tango
- Languages:
fluent in Mandarin Chinese and English
PUBLICATIONS
- Y. Zhao and G. Zhou,
“N-word-sequence Frequency Noise Mitigation for SLM Based on Binomial
Distribution”, Submitted to ICSLP2002: 2002 Intl. Conf. On Spoken Language
Processing, Denver, Sept. 2002.
- G. Zhou, M. Deisher,
and S. Sharma, “Causal Analysis of Speech Recognition Failure in Adverse
Environments”, Accepted to be published on IEEE ICASSP-02: 2002 Intl.
Conf. On Acoust., Speech, and Signal Processing, Orlando, FL. May, 2002.
- G. Zhou, J.H.L. Hansen,
and J.F. Kaiser, "Nonlinear Features for Classification and
Assessment of Speech Under Stress", IEEE Transactions on Speech and
Audio Processing, Vol. 9, No. 3, March 2001, pp. 201 - 216.
- J.H.L. Hansen, G. Zhou,
and Ruhi Sarikaya, "Analysis of Acoustic Correlates of Speech Under
Stress Part III: Applications to Speech Recognition, Speaker
Identification, and Stress Classification", accepted to Journal
Acoust. Society of America.
- G. Zhou, J.H.L. Hansen,
and J.F. Kaiser, "Methods for Stress Classification: Nonlinear TEO
and Linear Speech Based Features", Special Session of IEEE ICASSP-99:
1999 Intl. Conf. on Acoust., Speech, and Signal Processing, Phoenix, AZ,
Vol. 4, pp. 2087 - 2090, March 1999.
- G. Zhou, J.H.L. Hansen,
and J.F. Kaiser, "Linear and Nonlinear Speech Feature Analysis for
Stress Classification", ICSLP-98: 1998 Intl. Conf. on Spoken Lang.
Processing, Sydney, Australia, Vol. 2, pp. 883 - 886, Dec. 1998.
- G. Zhou, J.H.L. Hansen,
and J.F. Kaiser, "A New Nonlinear Feature for Stress
Classification", IEEE NORSIG-98, Denmark, pp. 89 - 92, June 1998.
- G. Zhou, J.H.L. Hansen,
and J.F. Kaiser, "Classification of
Speech Under Stress Based on Features Derived from the Nonlinear
Teager Energy Operator", IEEE ICASSP-98: 1998 Intl. Conf. on Acoust.,
Speech, and Signal Processing, Seattle, pp. 549 - 552, May 1998.
- G. Zhou, L. Zeng, and
C. Feng, "An Automatic Telephone Operator System Using Speech
Recognition", IEEE ICCT-96: 1996 Intl. Conf. on Comm. Technology,
Beijing, China, pp. 420 - 423, Oct. 1996.
- G. Zhou, L. Zeng, and
C. Feng, "The Performance Comparison of Conventional and Improved HMMs in the Speaker Independent
Speech Recognition of Isolated
Words", ICCT-94: 1994 Intl. Conf. on Comm. Technology, Shanghai,
China, pp. 1109 - 1112, Oct. 1994.