Towards Environment-Independent Spoken Language Systems

Abstract

In this paper we discuss recent results from our efforts to make SPHINX, the CMU continuous-speech speaker-independent recognition system, robust to changes in the environment. To deal with differences in noise level and spectral tilt between close-talking and desk-top microphones, we describe two novel methods based on additive corrections in the cepstral domain. In the first algorithm, an additive correction is imposed that depends on the instantaneous SNR of the signal. In the second technique, EM techniques are used to best match the cepstral vectors of the input utterances to the ensemble of codebook entries representing a standard acoustical ambience. Use of these algorithms dramatically improves recognition accuracy when the system is tested on a microphone other than the one on which it was trained.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1990
Accession Number
ADA457727

Entities

People

  • Alejandro Acero
  • Richard M. Stern

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Accuracy
  • Additives (Chemicals)
  • Algorithms
  • Automated Speech Recognition
  • Coefficients
  • Compensation
  • Computer Science
  • Computers
  • Databases
  • Degradation
  • Distortion
  • Environment
  • Frequency
  • Frequency Bands
  • Frequency Domain
  • Language
  • Noise

Fields of Study

  • Computer science

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Speech Processing/Speech Recognition.