Feature-Combination for Noise Robust Speech Pattern Processing

Research output

Research output per year 2007 2007 2008 2009 2009

• 5 Article

Research output per year

Research output per year

Search results

• 2009

  Improving automatic phoneme alignment under noisy conditions by incorporating spectral voicing information

  Jancovic, P. & Kokuer, M., 1 Jul 2009, In: Electronics Letters. 45, 14, p. 761-762 2 p.

  Research output: Contribution to journal › Article › peer-review

  • Additive noise 100%
  • Signal to noise ratio 70%
  • Experiments 30%
  1 Citation (Scopus)

• Incorporating the voicing information into HMM-based automatic speech recognition in noisy environments

  Jancovic, P. & Kokuer, M., May 2009, In: Speech Communication. 51, 5, p. 438-451 14 p.

  Research output: Contribution to journal › Article › peer-review

  • Automatic Speech Recognition 100%
  • Hidden Markov Model (HMM) 72%
  • Voicing 65%
  • Speech recognition 50%
  • Standard Model 19%
  12 Citations (Scopus)
• 2008

  Speech signal enhancement based on MAP algorithm in the ICA space

  Zou, X., Jancovic, P., Liu, J. & Kokuer, M., May 2008, In: IEEE Transactions on Signal Processing. 56, 5, p. 1812-1820 9 p.

  Research output: Contribution to journal › Article › peer-review

  • Independent component analysis 100%
  • Signal to noise ratio 18%
  35 Citations (Scopus)
• 2007

  Estimation of Voicing-Character of Speech Spectra Based on Spectral Shape

  Jancovic, P. & Kokuer Jancovic, M., Jan 2007, In: IEEE Signal Processing Letters. 14, 1, p. 66-69 4 p.

  Research output: Contribution to journal › Article › peer-review

  • Speech 100%
  • Character 58%
  • Spectrality 56%
  • False 30%
  • White noise 20%
  20 Citations (Scopus)

• Fast Algorithm for Calculation of the Union-Based Probability

  Jancovic, P., 1 Feb 2007, In: IEEE Transactions on Audio, Speech and Language Processing. 15, 2, p. 732-734 3 p.

  Research output: Contribution to journal › Article › peer-review