Text to speech syncronization

 

Alignement foresees the syncronization of each transcribed utterance in the textual resource with the corresponding acoustic signal and the simultaneous generation of the data bases of all utterances in the resource.

Utterance-based text-speech alignment of each corpus in C-ORAL-ROM has been realized through the speech software WPC with a method allowing the full apprecciation of both textual and acoustic information.

The text imported in an Align window and a tag ($)is inserted manually immediately after each terminal porsodic tag while the signal is plaied at a lower audio rate.

The program authomatically place the text on a speker layer aligned to the wave

 

 

Textual information can be fully exploit in C-ORAL-ROM format. The audio information can be accessed kliking on the text or in karaoke. The acoustic information is simultaneously displaied in the Transcribe window as in the following figure.