Speech recognition system, speech recognition result output method, and speech recognition result output program

• US 8,756,058 B2
• Filed: 02/22/2007
• Issued: 06/17/2014
• Est. Priority Date: 02/23/2006
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Adachi, Fumihiro
• Primary Examiner(s): He, Jialong
• Application Number:
  US12/280,253
• Publication Number:
  US 20100217598A1
• Time in Patent Office: 2,672 Days
• Field of Search: None
• US Class Current: 704/251
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/32 Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems
• CPC Class Codes: G10L 2015/027 Syllables being the recognition units

Abstract

A speech recognition system in which, even when the user makes an utterance including a word that satisfies a predetermined condition such as an unknown word, such a fact can be presented to the user, and the user can confirm the fact easily, is provided. The speech recognition system includes a word speech recognition section that converts input speech to a recognition result word sequence by using a predetermined word dictionary for recognition, a syllable recognition section that converts input speech to a recognition result syllable sequence, a segment determination section that determines a segment that corresponds to a predetermined condition which is a ground for estimating that a word in the converted recognition result word sequence is an unknown word, and an output section that obtains a partial syllable sequence from the recognition result syllable sequence corresponding to the determined segment, and outputs one or more word entries, which are in the vicinity of a position at which the partial syllable sequence is arranged in the word dictionary for recognition in which words are arranged in the order defined for word entries, together with the recognition result word sequence.

Claims

1. A speech recognition system, comprising:

a word speech recognition unit that converts input speech to a recognition result word sequence by using a predetermined word dictionary for recognition, the word dictionary including word entries arranged in a predetermined order;

a syllable recognition unit that converts the input speech to a recognition result syllable sequence;

a segment determination unit that determines a segment corresponding to an unknown word in the converted recognition result word sequence; and

an output unit that obtains a partial syllable sequence corresponding to the determined segment from the recognition result syllable sequence, determines an ordered position of the partial syllable sequence in the word dictionary on the assumption that the partial syllable sequence exists in the word dictionary, and displays on a display device one or more word entries in the vicinity of the ordered position in the word dictionary, together with the recognition result word sequence,

wherein the output unit displays the recognition result word sequence after replacing the recognition result word sequence with the partial syllable sequence with respect to the segment determined by the segment determination unit, and the output unit highlights the segment corresponding to the unknown word.

2. The speech recognition system according to claim 1, wherein the syllable recognition unit uses a predetermined syllable dictionary for recognition to convert the input speech to the recognition result syllable sequence.

3. The speech recognition system according to claim 1, wherein the predetermined order is the Japanese syllabary order.

4. The speech recognition system according to claim 1, wherein the output unit changes the output word entries of the word dictionary in accordance with an instruction of a user.

5. The speech recognition system according to claim 1, wherein when a user selects a word entry in the word dictionary, the output unit replaces a character sequence displayed in the segment determined by the segment determination unit with a character sequence used for display of the selected word entry.

6. The speech recognition system according to claim 1, wherein the output unit changes a range of the segment determined by the segment determination unit by an instruction of a user.

7. The speech recognition system according to claim 6, wherein the output unit changes an output content in accordance with a new segment obtained by changing the range in accordance with the instruction of the user.

8. A speech recognition result output method, comprising:

a word speech recognition step of converting input speech to a recognition result word sequence by using a predetermined word dictionary for recognition, the word dictionary including word entries arranged in a predetermined order;

a syllable recognition step of converting the input speech to a recognition result syllable sequence;

a segment determination step of determining a segment corresponding to an unknown word in the converted recognition result word sequence; and

an output step of obtaining a partial syllable sequence corresponding to the determined segment from the recognition result syllable sequence, determining an ordered position of the partial syllable sequence in the word dictionary on the assumption that the partial syllable sequence exists in the word dictionary, and displaying on a display device one or more word entries in the vicinity of the ordered position in the word dictionary, together with the recognition result word sequence,

wherein in the output step, the recognition result word sequence is displayed after the recognition result word sequence is replaced with the partial syllable sequence with respect to the segment determined by the segment determination step, and in the output step, the segment corresponding to the unknown word is highlighted.

9. The speech recognition result output method according to claim 8, wherein in the speech recognition step, a predetermined syllable dictionary for recognition is used to convert the input speech to the recognition result syllable sequence.

10. The speech recognition result output method according to claim 8, wherein the predetermined order is the Japanese syllabary order.

11. The speech recognition result output method according to claim 8, wherein in the output step, the displayed word entries of the word dictionary are changed in accordance with an instruction of a user.

12. The speech recognition result output method according to claim 8, wherein in the output step, when a user selects a word entry in the word dictionary, a character sequence displayed in the segment determined by the segment determination step is replaced with a character sequence used for display of the selected word entry.

13. The speech recognition result output method according to claim 8, wherein in the output step, a range of the segment determined by the segment determination step is changed in accordance with the instruction of a user.

14. The speech recognition result output method according to claim 13, wherein in the output step, an output content is changed in accordance with a new segment obtained by changing the range in accordance with the instruction of the user.

15. A non-transitory recording medium encoded with program code, when the program code is executed by a processor, performing:

word speech recognition processing for converting input speech to a recognition result word sequence by using a predetermined word dictionary for recognition, the word dictionary including word entries arranged in a predetermined order;

syllable recognition processing for converting the input speech to a recognition result syllable sequence;

segment determination processing for determining a segment corresponding to an unknown word in the converted recognition result word sequence; and

output processing for obtaining a partial syllable sequence corresponding to the determined segment from the recognition result syllable sequence, determining an ordered position of the partial syllable sequence in the word dictionary on the assumption that the partial syllable sequence exists in the word dictionary, and displaying on a display device one or more word entries in the vicinity of the ordered position in the word dictionary, together with the recognition result word sequence,

wherein in the output processing, the recognition result word sequence is displayed after the recognition result word sequence is replaced with the partial syllable sequence with respect to the segment determined by the segment determination processing, and in the output processing, the segment corresponding to the unknown word is highlighted.

16. The non-transitory recording medium according to claim 15, wherein in the syllable recognition processing, a predetermined syllable dictionary for recognition is used to convert the input speech to the recognition result syllable sequence.

17. The non-transitory recording medium according to claim 15, wherein the predetermined order is the Japanese syllabary order.

18. The non-transitory recording medium according to claim 15, wherein in the output processing, the displayed word entries of the word dictionary is changed in accordance with an instruction of a user.

19. The non-transitory recording medium according to claim 15, wherein in the output processing, when a user selects a word entry in the word dictionary for recognition, a character sequence displayed in the segment determined by the segment determination processing is replaced with a character sequence used for display of the selected word entry.

20. The non-transitory recording medium according to claim 15, wherein in the output processing, a range of the segment determined by the segment determination processing is changed in accordance with the instruction of a user.

21. The non-transitory recording medium according to claim 20, wherein an output content is changed in accordance with a new segment obtained by changing the range in accordance with the instruction of the user.