Browse Prior Art Database

Scheduled Simultaneous Voice Synthesis depending on significant parts in Text

IP.com Disclosure Number: IPCOM000013009D
Original Publication Date: 2001-Mar-01
Included in the Prior Art Database: 2003-Jun-12
Document File: 2 page(s) / 30K

Publishing Venue

IBM

Abstract

Disclosed is a system for improving to grasp the subject matter or to narrow it down efficiently in case of reading out several texts at once by using the text-to-speech (TTS) programs. Describes are the points of the invention as follows: 1. Sounding various kinds of vocal qualities and different rhythm allows the several contents to be read out simultaneously. (Simultaneous speech Synthesis) 2. Extracting the significant parts out of the texts, scheduling them not to clash with one another by which Simultaneous Speech Synthesis operates. (Simultaneous Utterance Scheduling)

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 1 of 2

Scheduled Simultaneous Voice Synthesis depending on significant parts in Text

Disclosed is a system for improving to grasp the subject matter or to narrow it down
efficiently in case of reading out several texts at once by using the text-to-speech (TTS) programs.

Describes are the points of the invention as follows:

1. Sounding various kinds of vocal qualities and different rhythm allows the several
contents to be read out simultaneously. (Simultaneous speech Synthesis)

2. Extracting the significant parts out of the texts, scheduling them not to clash with
one another by which Simultaneous Speech Synthesis operates. (Simultaneous
Utterance Scheduling)

3. The Simultaneous TTS program calculates the timing of the utterance and length of
pause as well as the speed of utterance, and it automatically selects the most suitable
schedule in order to sound the utterance without clashing.

4. In Simultaneous Speech Scheduling, the timing of utterance can be arranged to
sound at once when there is correspondence to read out more than one text. (e.g. An
English sentence to its Japanese translation)

5. Simultaneous Speech Scheduling enables to make use of electronic devices (like
stereo 3D speakers) efficiently by properly placing them in space. For example,
allowing more than one different synthesized speech sounds from each left and right
speaker differently.

6. When other speaking voices can be selected other than a male voice or a female voice,
from the perspective of the simultaneous speech synthesis, it automatically selects
the reader by which maximizes the distinction of voice quality and rhythm (speed,
pause, tone, characteristic of voice), and it also helps the listener to distinguish the
sounds.

For improving efficiency of reading out a text by TTS ,Fast-forward Speech Synthesis is
conventional technique:

  The fast-forward speech synthesis is a tool to improve efficiency of reading out a text
by using speech program. This system is to support the listener by skipping or
scanning the part(s) in which they find less important.

  However, it does not enable the listener to grasp the content of several texts at one
time. They are also not able to catch just a brief content and then carefully re-listen
for the content. Since texts are to be read out from the top to the end linearly, re-
listening, as the reader's voice has being fast-forwarded, tasks the listener indeed.

Disclosed method consists of following fun...