• Hynek Boril, Ph.D.


      Research Associate

      Center for Robust Speech Systems (CRSS)

      Erik Jonsson School of Engineering and Computer Science

      The University of Texas at Dallas

Research


Current Projects

Robust Speech Recognition in Adverse Environments

  • Data-driven front-end optimization
  • Unsupervised frequency/cepstral domain normalization of variations due to Lombard effect, noise, and channel
  • Improved back-end modeling of speech in noise utilizing codebooks of multi-environmental noisy models
  • Talking style/SNR/environment detection, improved scoring strategies in GMM-based multi-hypothesis testing
  • Mathematical modeling of noisy speech distributions in cepstral domain

Limited Resource Speech Recognition for Nigerian English

  • Study of acoustic-phonetic differences in Nigerian English (NE) and American English (AE)
  • Fast adaptation of AE based ASR to NE using limited resources

Speech Processing for In-Vehicle Systems (Part of UTDrive Project)

  • Driver's cognitive load/emotion classification
  • Assessment of speech dialog systems using multi-modal cognitive load analysis

Development of Infant Speech Production (Collaboration with LENA Foundation)

  • Longitudinal analysis of speech production parameters
  • Automatic age classification

Prosody Modeling

  • Statistical description of pitch contours utilizing codebook of pitch primitives
  • Application of pitch contour features to dialect distance assessment

Awards

  • 2006 - Principal Investigator (PI), “Normalization of Lombard Effect”, Siemens Corporate Technology (Munich, Germany), € 10,000 ($15,000), 1/1/2006 - 9/1/2006
  • 2006 - POSTER'06 - 10th International Student Conference on Electrical Engineering, Prague - awarded paper
  • 2005 - Interspeech'05 Student Travel Grant (ISCA)
  • 2005 - Comprehensive Doctoral Examination - Passed with Honors

Journal/Conference/Book Reviewer

  • Journal of the Acoustical Society of America (JASA)
  • EURASIP Journal on Audio, Speech, and Music Processing
  • IEEE International Conferences on Acoustics, Speech, and Signal Processing (ICASSP)
  • IEEE Workshops on Automatic Speech Recognition and Understanding (ASRU)
  • Book Reviews - Pavel Machac & Radek Skarnitzl (2009), Principles of Phonetic Segmentation, Epocha Publishing, Prague.

Professional Memberships (Past/Current)

  • IEEE
  • International Speech Communication Association (ISCA)
  • European Association for Signal Processing (EURASIP)
  • European Center of Excellence in Speech Synthesis (ECESS)

Past Projects

Projects & Evaluation Campaigns - Individual Participation

  • 2006 - Principal Investigator (PI), “Normalization of Lombard Effect”, Siemens Corporate Technology (Munich, Germany), € 10,000 ($15,000), 1/1/2006 - 9/1/2006
  • 2006 - European Center of Excellence on Speech Synthesis (ECESS) - First ECESS PDA/PMA Evaluation Campaign (2006) - Design and Evaluation of Pitch Tracker

European Union Projects - Research Group Member

  • 2006 - IST-2001-32216 - Lexica and Corpora for Speech-to-Speech Translation Components (LC-STAR II)
  • 2005 - COST 278 - Spoken Language Interaction in Telecommunication
  • 2003 - IST-1999-10003 - Speech-driven Interfaces for Consumer Devices (SPEECON)

Czech Science Foundation Grants - Research Group Member

  • 2005 - 2007 - GACR 102/05/0278 - New Trends in Research and Application of Voice Technology
  • 2004 - 2007 - 1ET201210402 - Voice Technologies in Information Systems
  • 2003 - 2006 - GACR 102/03/H085 - Biological and Speech Signals Modeling

Czech Government Projects - Research Group Member

  • 2005 - 2007 - MSM 6840770014 - Research in the Area of the Prospective Information and Navigation Technologies

Industry Projects - Group Member

  • 2005 - TEMIC SDS - Speech database processing
  • 2004 - TEMIC SDS - Acquisition of car speech database (CZKCC)

 

Last Updated 11.29. 2009