Multilingual speech recognition and public announcement

• US 9,015,032 B2
• Filed: 11/28/2011
• Issued: 04/21/2015
• Est. Priority Date: 11/28/2011
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Daye, Sheri G., Jaiswal, Peeyush, Vitenas, Aleksas J.
• Primary Examiner(s): ADESANYA, OLUJIMI A
• Application Number:
  US13/304,983
• Publication Number:
  US 20130138422A1
• Time in Patent Office: 1,240 Days
• Field of Search: 33/3, 33/8, 332/31, 332/35, 704/3, 704/8, 704/231, 704/235
• US Class Current: 704/8
• CPC Class Codes: G06F 9/454 Multi-language systems; Localisation; Internationalisation
• CPC Class Codes: G10L 13/027 Concept to speech synthesisers; Generation of natural phrases from machine-based concepts generation of parameters for speech synthesis out of text G10L13/08
• CPC Class Codes: G10L 15/005 Language recognition
• CPC Class Codes: G10L 25/54 for retrieval
• CPC Class Codes: H04R 27/00 Public address systems circuits for preventing acoustic reaction H04R3/02; circuits for distributing signals to loudspeakers H04R3/12; monitoring or testing arrangements for public address systems H04R29/007; amplifiers H03F

Abstract

Embodiments of the present invention provide a system, method, and program product to deliver an announcement to people, such as a public announcement. A computer receives input representative of audio from one or more people speaking in one or more natural languages. The computer processes the input to identify the languages being spoken, and identifies a relative proportion of each of the identified languages. Using these proportions, the computer determines one or more languages in which to deliver the announcement. The computer then causes to be delivered the announcement in the determined languages. In other embodiments, the computer can also determine an order in which to deliver the announcement. Further, the computer can transmit the announcement in the determined languages and order for delivery in aural or visual form.

Claims

1. A method for delivering an announcement in two or more languages, the method comprising the steps of:

a computer receiving, from a plurality of microphones deployed in an environment, input representative of audio of a group of human speakers speaking in the environment in two or more natural languages during a duration of time;

the computer processing the input to identify the two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

the computer processing the input to determine a relative proportion of each of the identified two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

the computer determining two or more natural languages in which to deliver the announcement based, at least in part, on the determined relative proportion of each of the identified two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

the computer determining a descending sequential order in which to deliver the announcement in the determined two or more natural languages based, at least in part, on the relative proportion of each of the identified two or more natural languages; and

the computer causing to be delivered the announcement in the determined two or more natural languages in the determined descending sequential order transmitting the announcement to an announcement system.

2. The method of claim 1, wherein the computer transmits for delivery the announcement in the determined two or more natural languages.

3. The method of claim 1, wherein:

the step of a computer receiving input further comprises;

a computer receiving input representative of audio from one or more human speakers speaking in two or more natural languages at set intervals of time; and

the step of the computer causing to be delivered the announcement further comprises;

the computer causing to be delivered the announcement in the determined two or more natural languages at set intervals of time.

4. The method of claim 1, wherein the announcement is caused to be delivered aurally.

5. The method of claim 1, wherein the announcement is caused to be delivered visually.

6. The method of claim 1, wherein the step of the computer determining two or more natural languages in which to deliver the announcement based, at least in part, on the determined relative proportion of each of the identified two or more natural languages comprises:

the computer determining a ranking of the identified two or more natural languages based, at least in part, on the determined relative proportion of each of the identified two or more natural languages; and

the computer determining two or more natural languages in which to deliver the announcement based, at least in part, on the determined ranking the identified two or more natural languages.

7. The method of claim 1, wherein the step of the computer causing to be delivered the announcement in the determined two or more natural languages in the determined descending sequential order comprises:

the computer identifying a specified delivery requirement, wherein the specified delivery requirement specifies a specific language to be delivered first, prior to the determined descending sequential order; and

the computer causing to be delivered the announcement in the determined two or more natural languages in the determined descending sequential order subsequent to the specific language of the specified delivery requirement.

8. The method of claim 1, wherein each microphone of the plurality of microphones transmits respective audio signals on separate channels.

9. The method of claim 1, wherein the plurality of microphones are positioned throughout a mass-transit passenger terminal.

10. The method of claim 9, wherein the computer processing the input to identify the two or more natural languages being spoken further comprises:

the computer accessing a model of an acoustical environment corresponding to positioning of the plurality of microphones.

11. The method of claim 1, wherein natural languages being spoken by more human speakers during the duration of time are determined to have a greater relative proportion than natural languages being spoken by fewer human speakers during the duration of time.

12. The method of claim 8, wherein the computer processing the input to determine a relative proportion of each of the identified two or more natural languages comprises:

the computer processing the audio signals on each channel to calculate a total number of utterance matches for each of the identified two or more natural languages across all channels; and

the computer ranking the identified two or more natural languages based on the calculated total number of utterance matches for each of the identified two or more natural languages.

13. A computer system for delivering an announcement in two or more languages, the system comprising:

a plurality of microphones deployed in an environment; and

one or more processors, one or more computer-readable memories, one or more computer-readable tangible storage devices, and program instructions stored on at least one of the one or more storage devices for execution by at least one of the one or more processors via at least one of the one or more memories, the program instructions comprising;

program instructions to receive, from the plurality of microphones deployed in the environment, input representative of audio of a group of human speakers speaking in the environment in two or more natural languages during a duration of time;

program instructions to process the input to identify the two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

program instructions to process the input to determine a relative proportion of each of the identified two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

program instructions to determine two or more natural languages in which to deliver the announcement based, at least in part, on the relative proportion of each of the identified two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

program instructions to determine a descending sequential order in which to deliver the announcement in the determined two or more natural languages based, at least in part, on the relative proportion of each of the identified two or more natural languages; and

program instructions to cause to be delivered the announcement in the determined two or more natural languages in the determined descending sequential order transmitting the announcement to an announcement system.

14. The system of claim 13, wherein the program instructions further comprise program instructions to receive and process the audio input at set intervals of time, and cause to be delivered the announcement in the determined two or more natural languages at set intervals of time.

15. The system of claim 13, wherein the program instructions further comprise program instructions to cause the announcement to be delivered aurally.

16. The system of claim 13, wherein the program instructions further comprise program instructions to cause the announcement to be delivered visually.

17. A computer program product for delivering an announcement, the computer program product comprising

one or more computer-readable tangible storage devices and program instructions stored on at least one of the one or more storage devices, the program instructions comprising:

program instructions to receive, from a plurality of microphones deployed in an environment, input representative of audio of a group of human speakers speaking in the environment in two or more natural languages during a duration of time;

program instructions to process the input to identify the two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

program instructions to process the input to determine a relative proportion of each of the identified two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

program instructions to determine two or more natural languages in which to deliver the announcement based, at least in part, on the relative proportion of each of the identified two or more natural languages being spoken in the environment by the group of human speakers during the duration of time;

program instructions to determine a descending sequential order in which to deliver the announcement in the determined two or more natural languages based, at least in part, on the relative proportion of each of the identified two or more natural languages; and

program instructions to cause to be delivered the announcement in the determined two or more natural languages in the determined descending sequential order transmitting the announcement to an announcement system.

18. The computer program product of claim 17, further comprising instructions, stored on at least one of the one or more storage devices, to receive and process the input at set intervals of time, and cause to be delivered the announcement in the determined two or more natural languages at set intervals of time.

19. The computer program product of claim 17, further comprising instructions, stored on at least one of the one or more storage devices, to cause the announcement to be delivered aurally.

20. The computer program product of claim 17, further comprising instructions, stored on at least one of the one or more storage devices, to cause the announcement to be delivered visually.