Signal Processing Electrical

Samuel Kim

**** *. ********** ***. *** ***, Los Angeles, CA 90089 310-***-****

********@***.*** http://worshipersam.net

Qualification/Skills

Research interest D igital signal processing and machine learning algorithms for multimedia data with applications

to information retrieval, compression, and enhancement: specifically audio, speech, and music

signals

Program languages C/C++, MATLAB, JAVA, PHP, HTML, XML, MySQL, Fortran

Tools Microsoft Visual Studio, KDevelop, Xcode, Eclipse, CMAKE, Subversion (SVN), WEKA, LaTex

Operating systems W indows, OS X, and Linux

E ducation

Ph.D., University of Southern California, Los Angeles, CA Dec. 2010

Electrical Engineering (Advisor: Dr. Shrikanth Narayanan)

Thesis t it le “Contextual modeling of audio signals toward information retrieval”

M .S., Yonsei University, Seoul, Korea Feb. 2005

Electrical and Electronic Engineering (Advisor: Dr. Hong-Goo Kang)

Thesis t it le “On feature extraction algorithms for GMM-based text independent speaker recognition systems”

B.S., Yonsei University, Seoul, Korea Feb. 2003

Electrical and Electronic Engineering (Advisor: Dr. Hong-Goo Kang)

P rofessional Experience

Research Assistant, University of Southern California, Los Angeles, CA 08/2005–present

Signal Analysis and Interpretation Lab. (SAIL)

• Speech interface system

o Developing speech recognition applications to various DARPA projects (along with Sphinx-based

O toSense)

o Multi-threaded, cross-platform, and real-time configurable programming based on C++ (utilizing

various libraries, such as BOOST, APR, LOG4CXX, PortAudio, PThread, etc.)

o Communicating through various message protocols, such as ActiveMQ and Open Agent Architecture

(OAA)

o Collaboration with BAE systems, Adapx, General Dynamics, and Institute of Creative Technology

( ICT)

• Microphone array system

o Real-time processing of audio signals from multiple microphones in SMARTROOM project using ASIO

o Speaker localization, speaker identification, audio segmentation, speech recognition, emotion

recognition

o Highly multi-thread and sophisticated synchronization with OpenCV-based video processing machines

• Music and audio information retrieval system

o Music fingerprint: compact context modeling of music with application to Classical music opus and

composer identification

o Acoustic topic models: contextual analysis of unstructured audio signals with application to audio

scene understanding

• Emotional speech analysis

o Analysis on emotional modulation of speech signals and glottal movement

o Real-time emotion recognition system (emotion mirror)

I n te rnship, Samsung Advanced I nstitute of Technology (SAI T), Kiheung, Korea 06/2005-08/2005

Human Computer Interaction (HCI) Lab.

• Music highlight extraction algorithms

Visiting Researcher, Chalmers University of Technology, Gothenburg, Sweden 07/2004-09/2004

Signals and Systems Department

• Speaker identification system; information theoretic performance analysis

Research Assistant, Yonsei University, Seoul, Korea 03/2003-06/2005

Media and Communication Signal Processing (MCSP) Lab.

• Speaker identification system

o Feature extraction algorithms for robust speaker identification systems

o Build a C++ speaker recognition library based on ITPP library

Teaching Assistant, Yonsei University, Seoul, Korea 03/2003-06/2004

Electrical and Electronic Engineering Department

• Random process, Information Theory, and Digital Signal Processing

Publications

1. Samuel Kim, Panayiotis G. Georgiou, and Shrikanth Narayanan, "Supervised acoustic topic model for

u nstructured audio information retrieval," in Asia Pacific Signal and Information Processing Association (APSIPA)

annual summit and conference, in press, Dec. 2010.

2. Samuel Kim, Shiva Sundaram, Panayiotis Georgious, and Shrikanth Narayanan, "An N-gram model for

u nstructured audio signals toward information retrieval," IEEE International Workshop on Multimedia Signal

P rocessing (MMSP), in press, Oct. 2010.

3. Daniel Bone, Samuel Kim, Sungbok Lee, and Shrikanth Narayanan, "A Study of Intra-Speaker and Inter-

Speaker Affective Variability using Electroglottograph and Inverse Filtered Glottal Waveforms," I NTERSPEECH,

i n press, Sep. 2010.

4. Samuel Kim, Shiva Sundaram, Panayiotis Georgiou, and Shrikanth Narayanan, "Context-based information

retrieval from unstructured audio signals," IEEE Transaction on Multimedia, in review, 2010.

5. Samuel Kim, Shiva Sundaram, Panayiotis Georgiou, and Shrikanth Narayanan, “Acoustic stopwords for

u nstructured audio information retrieval,” in European Signal Processing Conference (EUSIPCO), Aug. 2010.

6. Samuel Kim, Panayiotis Georgiou, Shrikanth Narayanan, and Shiva Sundaram, “Using naïve text queries for

robust audio information retrieval system,” IEEE International Conference on Acoustic, Speech and Signal

P rocessing (ICASSP), Mar. 2010.

7. Samuel Kim, Shiva Sundaram, Panayiotis Georgiou, and Shrikanth Narayanan, “Audio scene understanding

using topic models,” Neural Information Processing Systems (NIPS) Workshop on Applications for Topic Models:

Text and Beyond, Dec. 2009.

8. Samuel Kim, Shrikanth Narayanan, and Shiva Sundaram, “Acoustic Topic Model for Audio Information

Retrieval,” IEEE Workshop on Applications of Signal Processing to Audio and Acoustic (WASPAA), Oct. 2009.

9. Samuel Kim, Panayiotis G. Georgiou, and Shrikanth Narayanan, "A robust harmony structure modeling

scheme for Classical music opus identification," IEEE International conference on Acoustic, Speech and Signal

P rocessing (ICASSP), Apr. 2009.

10. Samuel Kim and Shrikanth Narayanan, "Dynamic chroma feature vectors with applications to cover song

i dentification," IEEE Multimedia Signal Processing (MMSP) Workshop, Oct. 2008.

11. C. Busso, M. Bulut, C. Lee, A. Kazemzadeh, E. Mower, S. Kim, J. Chang, S. Lee, and S. Narayanan,

" IEMOCAP: Interactive emotional dyadic motion capture database," Journal of Language Resources and

Evaluation, 2008.

12. Kyu Jeong Han, Samuel Kim, Shrikanth Narayanan, "Strategies to Improve the Robustness of Agglomerative

H ierarchical Clustering under Data Source Variation for Speaker Diarization," IEEE t ransaction of speech, audio,

and language processing, pp. 1590-1601, Nov. 2008.

13. Samuel Kim, Erdem Unal, and Shrikanth Narayanan, "Music fingerprint extraction for classical music cover

song identification," International Conference of Multimedia and Expo (ICME), pp. 1261-1264, May. 2008.

14. Kyu Jeong Han, Samuel Kim, and Shrikanth S. Narayanan, "Robust speaker clustering strategies to data

source variation for improved speaker diarization," Automatic Speech Recognition and Understanding (ASRU),

Dec. 2007.

15. Samuel Kim, Sungbok Lee, and Shrikanth Narayanan, "On voicing activity under the control of emotion and

loudness," 154th Meeting of the Acoustic Society of America (ASA), Nov. 2007.

16. Samuel Kim, Panayiotis G. Georgiou, Sungbok Lee, and Shrikanth Narayanan, “Real-time Emotion Detection

System using Speech: Multi-modal Fusion of Different Timescale Features,” Multimedia Signal Processing

( MMSP) Workshop, Oct. 2007.

17. Samuel Kim, Sungwan Yoon, Thomas Eriksson, Hong-Goo Kang, and Dae Hee Youn, “A noise-robust pitch

synchronous feature extraction algorithm for speaker recognition systems," Proc. I NTERSPEECH'05, p.p. 2029-

2032, Sep. 2005.

18. Thomas Eriksson, Samuel Kim, Hong-Goo Kang, and Chungyoung Lee, “An information-theoretic perspective

on feature selection in speaker recognition," IEEE Signal Processing Letters, Vol. 12, No. 7, pp. 500-503, Jul. 2005.

19. Samuel Kim, Jeong-Tae Seo, and Hong-Goo Kang, “Quantitative measure of speaker specific information in

h uman voice: from the perspective of information theoretic approach," Journals of Acoustical Society of Korea, vol.

24, Mar. 2005.

20. Samuel Kim, “On feature extraction algorithms for GMM-based text independent speaker recognition

systems," Master thesis, Yonsei University, Feb. 2005.

21. Sungwan Yoon, Samuel Kim, Hong-Goo Kang, and Dae Hee Youn, “An experiment of supplementary feature

for speaker verification in constrained environment," Proc. Korean Institute of Communication Sciences, vol.29,

p.p. 206, Nov. 2004.

22. Bongjin Lee, Samuel Kim, and Hong-Goo Kang, “Speaker recognition based on t ransformed line spectral

f requencies," Proc. International Symposium on Intelligent Signal Processing and Communication System, p.p.

177-180, Nov. 2004.

23. Samuel Kim, Thomas Eriksson, and Hong-Goo Kang, “On the t ime variability of vocal t ract for speaker

recognition," Proc. International Conf. Spoken Language Process. INTERSPEECH'04, vol. I I I, p.p. 2377-2380, Oct.

2004.

24. Thomas Eriksson, Samuel Kim, Hong-Goo Kang, and Chungyoung Lee, “Theory for speaker recognition over

IP," Proc. International Conf. Spoken Language Process. INTERSPEECH'04, vol. I, p.p. 625-628, Oct. 2004.

25. Sungwan Yoon, Samuel Kim, and Hong-Goo Kang, “Preliminary experiment searching for the supplementary

feature in speaker verification," Proc. ISISP, p.p.189-192, July 2004.

26. Samuel Kim, Thomas Eriksson, Hong-Goo Kang, and Dae Hee Youn, “A pitch synchronous feature extraction

method for speaker recognition," IEEE International Conference Acoustic Speech Signal Processing (ICASSP), vol.

I, p.p. 405-408, May 2004.

27. Thomas Eriksson, Samuel Kim, and Hong-Goo Kang, “On feature extraction in speaker recognition," Technical

Report 485L, Department of Signals and Systems, Chalmers University of Technology, Gothenburg, Sweden, Apr.

2004.

Patents

1. Method and system for pitch synchronous feature generation of speaker recognition system, Registered, KR 10-

052*****-***, 2005.

2. Method and apparatus for feature vector generation using a supplementary feature, Registered, KR 10-

056*****-***, March 2006.

3. Transformation method of speaker feature vector for speaker recognition, Registered, KR10-0596558-000, June

2006.

Services

Good Shepherd Ch ristian Group, University of Southern California, Los Angeles, CA 08/2005–present

Representative leader (01/2007 – 01/2008)

General leader (01/2006 – 08/2009)

Lamp P resbyterian Chu rch of Los Angeles, Gardena, CA 12/2005-present

Ordained deacon (11/2009 – present)

Music director (03/2007 – present)

Tsunami Relief Activity, T r incomollee, Sri Lanka 03/2005

Youth With A M ission 09/2000 – 07/2001

Jabalpur, MP, India / Charlotte, NC / Los Angeles, CA

M i lita ry Service 03/1998 – 05/2000

Korean Army, Wonju, Kangwondo, Korea

Contact this candidate