MAN/MACHINE COMMUNICATIONS SYSTEM USING FORMANT BASED SPEECH ANALYSIS AND SYNTHESIS

摘要

MAN/MACHINE COMMUNICATIONS SYSTEM USING FORMANT BASED SPEECHANALYSIS AND SYNTHESISABSTRACT OF THE DISCLOSUREFormants are extracted and stored from reference speech.Input speech is suitably processed to derive unlabelled candidateformants. The sets of formants from the input and referencespeech are compared using dynamic programming techniques. Afurther sequence comparison provides time alignment of the inputand reference speech. The sequence comparisons extract a dissim-ilarity measure based on the formant frequencies and othercharacteristics of the speech. The reference speech resulting inthe lowest dissimilarity measure identifies the input speechrecognized by the system. System feedback may be provided and iscomposed of designated responsive multi-voiced speech. Themulti-voiced output speech is obtained primarily by altering theprosodic parameters and formant frequencies of the designatedresponsive speech. Thus, the designated responsive speech may,say in an aircraft communication system, use one voice outputwhen providing an information response to the pilot's recognizedinput speech question and another appropriately strident voice toissue the pilot warnings. The system also may be placed in atraining mode to evaluate performance and adjust parameters.