Utilizing word-to-digital conversion

• US 4,707,858 A
• Filed: 05/02/1983
• Issued: 11/17/1987
• Est. Priority Date: 05/02/1983
• Status: Expired due to Term

• Current Assignee: Motorola, Inc. (Motorola Solutions, Inc.)
• Original Assignee: Motorola, Inc. (Motorola Solutions, Inc.)
• Inventors: Fette, Bruce A.
• Primary Examiner(s): Shaw, Gareth D.
• Assistant Examiner(s): Mills, John G.
• Application Number:
  US06/490,701
• Time in Patent Office: 1,660 Days
• Field of Search: 381/29-53
• US Class Current: 704/251
• CPC Class Codes: G10L 19/0018 Speech coding using phonetic or linguistical decoding of the source; Reconstruction using text-to-speech synthesis

Abstract

A communications system each end of which includes means for analyzing human speech and comparing each word to prestored words for word and speacker recognition, the message then being digitized along with characteristic properties of the speackers voice to form a signal for transmission having a rate of approximately 75 bits per second, transmitting the digitized message to a remote terminal which converts it to a spoken message in the synthesized voice of the original speaker.

Claims

1. A method of extremely narrowband communication comprising the steps of:

converting human speech to electrical signals;

analyzing the electrical signals to provide a plurality of signals representative of a plurality of properties which characterize a human voice;

storing signals representative of a plurality of spoken words;

comparing at least some of the plurality of signals to the stored signals to determine specific words in the human speech and supplying signals representative of the specific words; and

converting the supplied signals representative of specific words to a digital form having a rate of less than 300 bits per second.

2. A method as claimed in claim 1 including the step of recognizing the beginning and the end of each spoken word prior to the step of comparing.

3. A method as claimed in claim 2 including in the storing step, storing signals representative of a plurality of words spoken by a plurality of different individuals and further including in the comparing step the supplying of signals representative of the individual speaking the specific words.

4. A method as claimed in claim 2 including the steps of storing a plurality of predetermined messages and indicating to the speaker a list of possible next words subsequent to the recognition of the end of a word.

5. A method as claimed in claim 3 including in addition the steps of formatting the human speech, after conversion to digital form, into a digital electrical signal containing a plurality of bits representative of the message and a plurality of bits representative of characteristic properties of the human voice and transmitting the digital electrical signal to a remote terminal.

6. A method as claimed in claim 5 including the steps of receiving a digital electrical signal transmitted from a remote terminal and coverting the received signal to a spoken message in a synthesized voice having approximately the characteristic properties of an original speaker at the remote terminal.