Apparatus, method and computer program product for translating speech, and terminal that outputs translated speech

1. A speech translation apparatus comprising:

a correspondence storage unit that stores identifiers to uniquely identify each of a plurality of terminals connectable via a network and usage languages used in the terminals associated with each other;

a receiving unit that receives a source speech from one of the terminals;

a translating unit that acquires the usage languages different from a source language used in the source speech from the correspondence storage unit, and generates a translated speech translated from the source speech by using each of the acquired usage languages as a target language;

a determining unit that determines whether the translated speech has been generated in all of the target languages when the translated speech is generated in any one of the target languages;

an output unit that outputs the translated speech when the translated speech has been generated in all of the target languages; and

a sending unit that sends the translated speech to each of the terminals identified by the identifier corresponding to the target language.

2. The apparatus according to claim 1 , wherein

the output unit further outputs the source speech when the translated speech has been generated in all of the target languages, and

the sending unit sends the source speech to the terminal identified by the identifier corresponding to the source language.

3. The apparatus according to claim 1 , wherein

the determining unit determines whether a first speech duration and a second speech duration of the translated language have been generated in all of the target languages, when the translated speech is generated in any one of the target languages, the first speech duration being a duration to be output next in one of the target languages and the second speech duration being a duration to be output in any other target language before an end of the first speech duration, and

the output unit outputs the translated speech and the source speech that correspond to the first speech duration and the second speech duration, when the first speech duration and a second speech duration of the translated speech have been generated in all of the target languages.

4. The apparatus according to claim 1 , wherein the translating unit generates the translated speech a reproduction time of which lasts for a substantially same length of time as that in the target languages.

5. The apparatus according to claim 4 , wherein the translating unit changes reproduction speed of the translated speech to make the length of the translated speech substantially same among a plurality of the target languages, when the reproduction time of the translated speech does not substantially same as that of the target languages.

6. The apparatus according to claim 4 , wherein the translating unit adds a silent speech at least one of before and after the translated speech to make the reproduction time of the translated speech substantially same among a plurality of the target languages, when the reproduction time of the translated speech does not substantially same as that of the target languages.

7. The apparatus according to claim 1 , further comprising a speech storage unit that is capable of storing a mixed speech obtained by mixing the translated speeches or the source speeches with respect to each target language, wherein

the output unit further mixes the mixed speech stored in the speech storage unit with an additional translated speech or another source speech with respect to each target language, stores the mixed speech mixed by the output unit in the speech storage unit, and outputs the mixed speech when the translated speech is generated in all of the target languages.

8. The apparatus according to claim 1 , further comprising a speech storage unit that is capable of storing the translated speech or the source speech in each target language with respect to each terminal, wherein

the output unit stores the translated speech and the source speech in the speech storage units and outputs the translated speech and the source speech in the target languages acquired from the speech storage unit, when the translated speech has been generated in all of the target languages, and

the sending unit sends a mixed speech including only the translated speeches to the terminal that sent the source speech, and sends a mixed speech including the translated speech and the source speech to other terminals.

9. The apparatus according to claim 1 , further comprising a delay unit that computes a first difference by subtracting a second time point at which the source speech is received from a first time point at which the translated speech is determined to have been generated in all of the target languages, and delays output of the translated speech and the source speech until the first threshold time has passed since the second time point, when the computed first difference is less than a first threshold time determined in advance, wherein

the output unit outputs the translated speech and the source speech after the delay unit delays the reproduction.

10. The apparatus according to claim 9 , further comprising a shortening unit that shortens a reproduction time of the translated speech and the source speech by a length of a second difference computed by subtracting the first threshold time from the first difference, when the first difference is more than the first threshold time.

11. The apparatus according to claim 10 , wherein the shortening unit shortens the reproduction time of the translated speech and the source speech by increasing a reproduction speed.

12. The apparatus according to claim 10 , wherein the shortening unit detects at least one of a silence and a noise from each of the translated speech and the source speech and shortens the reproduction time of the translated speech and the source speech by eliminating the silence and the noise.

13. The apparatus according to claim 9 , wherein the delay unit computes the first difference for each of a predetermined number of second speech durations of which the translated speech and the source speech have been output before a first speech duration of which the translated speech has been generated in all of the target languages, computes a product of an average of first differences and a predetermined coefficient, and delays reproduction of the first speech duration of the translated speech and the source speech until the first threshold time has passed since the second time point, when the product is less than the first difference.

14. The apparatus according to claim 13 , wherein the delay unit computes the product using the predetermined number of the second speech durations of which the translated speech and the source speech have been output before the first speech duration, each of the second durations being more than a second threshold time determined in advance.

15. The apparatus according to claim 1 , wherein

the receiving unit further receives an image associated with the source speech from the terminal, and

the sending unit sends the image further associated with the translated speech or the source speech.

16. The apparatus according to claim 15 , wherein the translating unit generates the translated speech a reproduction time of which is substantially same as that in the source speech.

17. The apparatus according to claim 16 , wherein the translating unit generates the translated speech a reproduction time of which is substantially same as that of the source speech by adding a silent speech at least one of before and after the translated speech, when the reproduction time of the translated speech is shorter than that of the source speech.

18. A speech translation apparatus comprising:

a correspondence storage unit that stores identifiers to uniquely identify each of a plurality of terminals connected via a network and usage languages used in the terminals associated with each other;

a receiving unit that receives a source speech from one of the terminals;

a translating unit that acquires the usage languages different from a source language used in the source speech from the correspondence storage unit, and generates a translated speech translated from the source speech by using each of the acquired usage languages as a target language;

a sending unit that sends the translated speech to each of the terminals identified by the identifier corresponding to the target language and sends the source speech to each of the terminals identified by the identifier corresponding to the source language;

a determining unit that determines whether the translated speech has been generated in all of the target languages, when the translated speech is generated in any one of the target languages; and

an output unit that outputs duration information of a speech duration of the source speech determined by the determining unit, when the translated speech has been generated in all of the target languages, wherein

the sending unit further sends the duration information to the terminals.

19. A speech translation apparatus comprising:

a language storage unit that stores usage languages;

a first receiving unit that receives a source speech from a plurality of other speech translation apparatuses connectable via a network;

a translating unit that generates a translated speech translated from the source speech by using each of the usage languages stored in the language storage unit as a target language;

a second receiving unit that receives information indicating that the source speech has been translated into a language used in another speech translation apparatus therefrom;

a determining unit that determines whether the translated speech has been generated in all of the other speech translation apparatuses, when the information is received from any one of the other speech translation apparatuses; and

an output unit that outputs the translated speech when the translated speech has been generated in all of the other speech translation apparatuses.

20. A speech translation method comprising:

receiving a source speech from a plurality of terminals connectable via a network;

acquiring the usage languages different from a source language used in the source speech from a correspondence storage unit that stores identifiers to uniquely identify each of a plurality of terminals and usage languages used in the terminals associated with each other;

generating a translated speech translated from the source speech by using each of the usage languages acquired as a target language;

determining whether the translated speech has been generated in all of the target languages when the translated speech is generated in any one of the target languages;

outputting the translated speech when it is determined that the translated speech has been generated in all of the target languages; and

sending the translated speech to each of the terminals identified by the identifier corresponding to the target language.

21. A computer program product having a computer readable medium including programmed instructions for translating a source speech, wherein the instructions, when executed by a computer, cause the computer to perform:

receiving the source speech from a plurality of terminals connectable via a network;

acquiring the usage languages different from a source language used in the source speech from a correspondence storage unit that stores identifiers to uniquely identify each of a plurality of terminals and usage languages used in the terminals associated with each other;

generating a translated speech translated from the source speech by using each of the usage languages acquired as a target language;

determining whether the translated speech has been generated in all of the target languages when the translated speech is generated in any one of the target languages;

outputting the translated speech when it is determined that the translated speech has been generated in all of the target languages; and

sending the translated speech to each of the terminals identified by the identifier corresponding to the target language.