• DocumentCode
    2268839
  • Title

    Improving automatic speech recognition robustness for the Romanian language

  • Author

    Buzo, Andi ; Cucu, Horia ; Burileanu, Corneliu

  • Author_Institution
    Fac. of Electron., Telecommun. & Inf. Technol., Univ. “Politeh.” of Bucharest, Bucharest, Romania
  • fYear
    2011
  • fDate
    Aug. 29 2011-Sept. 2 2011
  • Firstpage
    2119
  • Lastpage
    2122
  • Abstract
    In this paper we propose an alternative way of improving speech recognition accuracy by analyzing the relevance of voice feature dimensions. A new measure is defined in order to quantify the feature distribution overlapping. Based on this measure, weights for voice feature dimensions are calculated and then applied to the hypotheses resulted from an N-best recognition process. Experiments are made with an Automatic Speech Recognition (ASR) system for the Romanian language. A relative improvement of 22% is obtained in terms of Word Error Rate (WER).
  • Keywords
    feature extraction; natural language processing; speech enhancement; speech recognition; ASR system; N-best recognition process; Romanian language; WER; automatic speech recognition robustness improvement; voice feature dimensions; voice feature distribution overlapping; word error rate; Accuracy; Databases; Equations; Hidden Markov models; Mathematical model; Speech recognition; Uncertainty;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Signal Processing Conference, 2011 19th European
  • Conference_Location
    Barcelona
  • ISSN
    2076-1465
  • Type

    conf

  • Filename
    7074079