Real-time transcription correction system

• US 20030212547A1
• Filed: 05/13/2003
• Published: 11/13/2003
• Est. Priority Date: 04/08/1999
• Status: Active Grant

• Current Assignee: Ultratech Incorporated (Veeco Instruments Incorporated)
• Original Assignee: Ultratech Incorporated (Veeco Instruments Incorporated)
• Inventors: Vitek, Troy D., Colwell, Kevin R., Engelke, Robert M., Turner, Jayne M., Frazier, Pamela A., Grittner, Kurt M.
• Granted Patent: US 7,164,753 B2
• Time in Patent Office: Days
• US Class Current: 704/200
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/225 Feedback of the input speech
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 2201/60 Medium conversion
• CPC Class Codes: H04M 3/42391 where the subscribers are hearing-impaired persons, e.g. telephone devices for the deaf

Abstract

A voice transcription system employing a speech engine to transcribe spoken words, detects the spelled entry of words via keyboard or voice to invoke a database of common words attempting to complete the word before all the letters have been input. This database is separate from the database of words used by the speech engine. A voice level indicator is presented to the operator to help the operator keep his or her voice in the ideal range of the speech engine.

Claims

1. A voice transcription system comprising:

an input circuit receiving a first voice signal of spoken words from a remote source;

a speech engine receiving a second voice signal to generate input text including first text words corresponding to the spoken words;

a monitor circuit receiving the second voice signal and producing a speech level signal related to the strength of the second voice signal;

a display device viewable by the call assistant displaying the input text and displaying the speech level signal as a graphic; and

an output circuit transmitting the input text to a remote location.

2. The voice transcription system of claim 1 wherein the first and second voice signals are the same.

3. The voice transcription system of claim 1 wherein the second voice signal is a revoicing of the first voice signal by the call assistant.

4. The voice transcription system of claim 1 wherein the graphic is a constant area shape whose color indicates speech level.

5. The voice transcription system of claim 4 wherein the color is one of a set of three colors indicating low volume, proper volume and excessive volume of the second voice signal.

6. The voice transcription system of claim 5 wherein the color associated with low volume is a background screen color, the color associated with proper volume is green and the color associated with excessive volume is red.

7. The voice transcription system of claim 5 wherein at least one color includes a visual texture distinguishing the color for call assistants who have color blindness.

8. The voice transcription system of claim 1 wherein the call assistant voice level signal is a function of average power of the revoice signal over a predetermined interval.

9. The voice transcription system of claim 8 wherein the speech level signal is a function evaluating three ranges into which the average power falls, the ranges defining:

insufficient power, proper power, and excessive power for the speech engine.

10. A voice transcription system comprising:

an input circuit receiving a voice signal of spoken words from a remote source;

a speech engine generating input text including first text words corresponding to the spoken words;

a call assistant input circuit receiving at least one letter input for second text words from a call assistant monitoring at least one of the voice signal and the input text, the call assistant input circuit including a database of second text words to select a second text word from the letter input; and

an output circuit transmitting at least one of the first text words and the second text word to a remote user.

11. The voice transcription system of claim 10 further including a display device viewable by the call assistant and having a screen displaying the input text as printed text;

and

wherein the second text word is displayed to the call assistant on the display device and wherein the call assistant input circuit includes a means for the call assistant to accept or reject the displayed second text word.

14. The voice transcription system of claim 11 wherein the display device provides a touch screen identifying a point of touch on the screen and wherein the displayed second word is accepted by a touch of the screen at a location of the displayed second word.

15. The voice transcription system of claim 11 wherein the displayed second word is accepted by entry of a designated entry character.

16. The voice transcription system of claim 15 wherein the designated entry character is an enter key on a keyboard.

17. The voice transcription system of claim 10 wherein the call assistant input circuit accepts typed letter input from the call assistant via a keyboard.

18. The voice transcription system of claim 10 wherein the call assistant input circuit accepts spoken letter input from the call assistant via the speech engine.

19. The voice transcription system of claim 10 wherein the memory includes multiple sets of second words and wherein the input circuit further receives a remote source identifier to select from among the sets of second words for selecting a second word.

20. The voice transcription system of claim 10 wherein the remote identifier is a caller identification number.

21. The voice transcription system of claim 10 wherein the remote identifier is an electronically readable serial number of a telephone device.

12. The voice transcription system of claim 111 wherein the displayed second word is rejected upon receipt by the call assistant input circuit receiving an additional letter input not conforming to the displayed second text word.

13. The voice transcription system of claim 12 wherein upon a rejection of the displayed second word by receipt of an additional letter input not conforming to the displayed second text word, the call assistant input circuit selects a new second text word from the database.