I recently joined IBM India Research Lab, New Delhi, India as a Research Staff Member.

I successfully completed my Ph.D. defense on August 25th 2006! ( Abstract ) ( Thesis ).

Education

• 2006: Ph.D., Electrical & Computer Engineering, University of Maryland, College Park, USA.
• 2001: M.S., Electrical & Computer Engineering, Boston University, Boston, MA, USA
• 1999: B.E. (Hons.), Electrical & Electronics Engineering, Birla Institute of Technology & Science, Pilani, India

Research Projects

• Speech Enhancement and Robust Speech Recognition

  In this project, we are developing algorithms for speech enhancement and robust speech recognition using the auditorily motivated Modified Phase Opponency (MPO) model and the Aperiodicity, Periodicity and Pitch (APP) detector. The goal of this project is to develop a speech enhancement technique that will:

  • Efficiently enhance speech signals corrupted by additive colored noise with fluctuating levels.
  • Introduce little or no distortion to clean speech signals.
  • Not need an estimate of the noise characteristics.
  • Improve the performance of Automatic Speech Recognition (ASR) in noisy conditions.

• Detection of Aperiodicity & Periodicity in speech signals

  In this work, we developed an algorithm, called the Aperiodicity, Periodicity and Pitch (APP) detector, to compute direct measures of aperidoicity and periodicity in a speech signal using temporal information. These measures are a part of the temporal based landmark detection project. The APP detector has applications in a wide range of area including speaker recognition, voice quality evaluation, accent detection and speech recognition.

• Knowledge based Acoustic Parameters vs MFCCs

  This project involved the evaluation of the knowledge based acoustic parameters developed in our lab as front end for speech recognition and comparing the performance of these parameters with that of the Mel Frequency Cepstral Coefficients (MFCCs). We showed that the APs are more robust to linear filtering variations than the MFCCs.

Awards and Honors

1. Royster Student Scholarship Award for Best Poster at the Joint Meeting of the Acoustical Society of America's North Carolina and Washington DC chapters, 2005
2. Best Student Paper Award in Speech Communication at the Acoustical Society of America Meeting, 2003
3. Jacob K. Goldhaber Travel Grant, University of Maryland, 2003.
4. Best Student Paper Award in Speech Communication at the Acoustical Society of America Meeting, 2001
5. Best 'Simulation-Model Exhibit' Award at A Professions-Oriented Gathering Over Educational Experiences(APOGEE), BITS, Pilani, 1998

Publications

• Invited Talks
  1. "Synergy of Acoustic-Phonetics and Auditory Modeling towards Speech Enhancement and Robust Speech Recognition Enhancement Using Modified Phase Opponency Model", Hearing Seminar at the Center for Computer Research in Music and Acoustics, Stanford University, Nov. 2006.
  2. "Speech Enhancement Using Modified Phase Opponency Model", Joint meeting of Acoustical Society of America's North Carolina and Washington DC chapters, Hampton, VA, Nov. 2005.
• Journal Papers
  1. "Speech Enhancement Using The Modified Phase Opponency Model", in print, Journal of Acoustical Society of America.
  2. "The Development and Testing of a Phase-Opponent Noise-Reduction Algorithm", in revision EURASIP Journal on Applied Signal Processing.
  3. Om Deshmukh, Carol Espy-Wilson, Ariel Salomon and Jawahar Singh, "Use of Temporal Information: Detection of the Periodicity and Aperiodicity Profile of Speech", IEEE Transactions on Speech and Audio Processing, Vol. 13 (5), pp. 776-786, Sept. 2005, (paper pdf ).
  4. Ariel Salomon, Carol Espy-Wilson and Om Deshmukh, "Detection of Speech Landmarks:  Use of Temporal Information", J. Acoust. Soc. Am., vol. 115, pp. 1296-1305, March 2004, ( paper pdf ) .
• Conference Papers
  1. Om Deshmukh, Carol Espy-Wilsin, "Evaluating the perceptual quality of speech signals enhanced using the Modified Phase Opponency model", 152nd meeting of the Acoustical Society of America, Hawaii, 2006.
  2. Om Deshmukh, Carol Espy-Wilson, "Speech Enhancement Using Modified Phase Opponency Model", Proceedings of Interspeech 2006, Pittsburgh, pp. 269-272.(pdf) (talk).
  3. Om Deshmukh, Carol Espy-Wilson, "Modified Phase Opponency Based Solution to the Speech Separation Challenge", Proceedings of Interspeech 2006, Pittsburgh, pp. 101-104. (pdf) (talk).
  4. Om Deshmukh, Carol Espy-Wilson, "Speech Enhancement based on Modified Phase-Opponency Detectors", 150th meeting of Acoustical Society of America, Minneapolis, Minnesota, 2005, (poster ppt).
  5. Om Deshmukh, Carol Espy-Wilson, "Speech Enhancement Using Auditory Phase Opponency Model", Proceedings of Eurospeech, Lisbon, Portugal, pp. 2117-2120, 2005, (paper pdf), (poster ppt)
  6. Om Deshmukh, Michael Anzalone, Carol Espy-Wilson, Laurel Carney, "A Noise-Reduction Strategy for Speech based on Phase-Opponency Detectors", 149th meeting of Acoustical Society of America, Vancouver Canada, 2005, (poster ppt), (poster pdf).
  7. Om Deshmukh, Carol Espy-Wilson, "A Novel Method for Computation of Periodicity, Aperiodicity and Pitch of Speech Signals", in Proceedings of IEEE ICASSP, 2004, Montreal, Canada, pp. I.117-120(paper pdf,(talk ppt)
  8. Om Deshmukh, Carol Espy-Wilson, "Detection of Periodicity and Aperiodicity in Speech Signal Based on Temporal Information ", The 15th International Congress of Phonetic Sciences, Barcelona, Spain, pp. 1365-1368, 2003. (paper pdf).
  9. [Best presentation award] Om Deshmukh, Carol Espy-Wilson, "A measure of Aperiodicity content in Speech", 145th meeting of Acoustical Society of America, Nashville, TN, 2003. (talk ppt),(Abstract)
  10. Om Deshmukh, Carol Espy-Wilson, "A measure of Periodicity and Aperiodicity in Speech", in Proc. IEEE ICASSP 2003, Hong Kong, pp. 448-451. (paper pdf)
      NOTE:This conference was cancelled due to SARS.
  11. Om Deshmukh, Espy-Wilson, C. and Juneja, A., "Acoustic-phonetic speech parameters for speaker-independent speech recognition", in Proc. IEEE ICASSP 2002, May 13-17, 2002, Orlando, Florida, pp. 593-596 (paper pdf), (poster pdf)
  12. Beth Logan, Pedro Moreno and Om Deshmukh, "Word and Sub-word Indexing Approaches for Reducing the Effects of OOV Queries on Spoken Audio", Human Language Technology Conference (HLT), March 2002, (paper pdf)
  13. Amit Juneja, Om Deshmukh and Carol Espy-Wilson, "An Event-Based Acoustic-Phonetic Approach For Speech Segmentation And E-set Recognition", Presented in the student Forum at ICASSP 2002, (poster ppt)
  14. [Best presentation award]Om Deshmukh, Carol Espy-Wilson, Ariel Salomon "Robust speech event detection using strictly temporal information.", 141st meeting of ASA, Chicago, IL, 2001. (abstract)

TECH ( This event was previously known as Research Review Day) Publications