융합신호처리학회논문지 (Journal of the Institute of Convergence Signal Processing)

한국융합신호처리학회 (The Korea Institute of Convergence Signal Processing)
권호 표지

에코제거기와 MAP 추정에 기초한 핸즈프리 음성 인식

Hands-free Speech Recognition based on Echo Canceller and MAP Estimation

  • 발행 : 2003.07.01
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

핸즈프리 마이크를 이용한 원격회의나 원격 통신 시스템과 같은 몇 가지의 응용분야에서, 음성 신호는 주위 잡음뿐만 아니라 마이크와 스피커사이의 결합에 의해 발생하는 에코에 의해서 왜곡되기 쉽다. 게다가 채널 왜곡이나 부가적인 잡음을 포함한 환경 잡음들은 원래의 입력 음성신호에 영향을 미치리라 고려된다. 본 논문에서는, 이러한 핸즈프리 음성에 있어서의 음성 인식률을 향상시키기 위해 에코 제거기와 최대 사후 추정(MAP)을 이용한 새로운 접근방식을 소개한다. 이 접근방식에서, 제안된 시스템이 에코를 포함한 주위 잡음 환경에서의 핸즈프리 음성인식에 효과적이라는 것을 보여준다 또한, 실험 결과는 에코 제거기와 MAP 환경적응 기술의 결합 시스템이 에코와 잡음 환경에 잘 적응하는 것을 보여준다.

For some applications such as teleconference or telecommunication systems using a distant-talking hands-free microphone, the near-end speech signals to be transmitted is disturbed by an ambient noise and by an echo which is due to the coupling between the microphone and the loudspeaker. Furthermore, the environmental noise including channel distortion or additive noise is assumed to affect the original input speech. In the present paper, a new approach using echo canceller and maximum a posteriori(MAP) estimation is introduced to improve the accuracy of hands-free speech recognition. In this approach, it was shown that the proposed system was effective for hands-free speech recognition in ambient noise environment including echo. The experimental results also showed that the combination system between echo canceller and MAP environmental adaptation technique were well adapted to echo and noise environment.

키워드

참고문헌

  1. Proc. ICASSP Hands Free Continuous Speech Recognition in Noisy Environment using a Four Microphone Array D.Giuliani.;M.Matassoni;M.Omologo;P.Svaizer
  2. Proc. CASSP Model Adaptation based on HMM Decomposition for Reverberant Speech Recognition T.Takiguchi;S.Nakamura;Q.Huo;K.Shikano
  3. Proc. of ICASSP v.1 Training of HMM with Filtered Speech Material for Hands-free Recognition D.Giuliani;M.Matassoni;M.Omologo;P.Svaizer
  4. Robust Methods for Speech Recognition in Adverse Conditions(ROBUST-99) Multi-microphone noise reduction techniques for handsfree speech recognition - a comparative study J.Bitzer;K.U.Simmer;K.Kammeyer
  5. Proc. ICASSP Coupled Adaptive Filters for Acoustic Echo Control and Noise Reduction R.Martin;J.Altenhoner
  6. Proc. IEEE ICASSP New Techniques for Adaptive Filtering Applied to Speech Echo Cancellation M.Siqueira;A.Alwan
  7. Proc. ICASSP'90 Acoustic echo canceller algorithm based on the variation characteristics of a room impulse response S.Makino;Y.Kaneda
  8. Proc. ICASSP Analysis and Design of Multirate Systems for Cancellation of Acoustical Echoes W.Kellerman
  9. Proc. ICSLP An Unsupervised Speaker Adaptation Method for Continuous Parameter HMM by Maximum A Posteriori Probability Estimation Y.Tsurumi;S.Nakagawa
  10. Computer Speech and Language v.9 Maximum likelihood linear regression for speaker adaptation of continuous density hidden Markov models C.J.Leggetter;P.C.Woodland
  11. Proc. European Conference on Speech Communication and Technology v.1 Maximum a posteriori linear regression for hidden Markov model adaptation C.Chesta;O.Siohan;C.-H.Lee
  12. Proc. IEEE Invited Workshop on Pattern Recognition for Multimedia Techniques Environmen- tal Adaptation using A Posteriori Estimation for Korean Word Recognition J.H.Lee;B.K.Kim;H.Y.Chung