JP3106550B2 - Voice recognition result display device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a speech recognition result display device.

[0002]

2. Description of the Related Art Conventionally, when a speech recognition result is displayed, either a kana-kanji candidate or a kana-kana candidate is displayed. For example, if “Hashiwa” is uttered and displayed with kana-kanji mixed candidates, “
At the end of the bridge, chopsticks are at the end of chopsticks, and the name of the ashi is ‥ ”. When displayed as a candidate for the reading kana, it becomes“ Ashiha, Hashiha, Hashishima, Hajiha, Ajiha, ‥ ”.

[0003]

However, when displaying with kana-kanji mixed candidates, there are homophones even with the same reading, and a plurality of candidates are created due to differences in how words are cut. The number of candidates to be used becomes enormous. Therefore, it is good if the correct answer candidate is displayed at the top, but at other times, the correct answer candidate is confused with a huge number of incorrect recognition / incorrect conversion candidates, and it takes time to search for the correct answer candidate.

[0004] On the other hand, in the case of displaying a candidate of a reading kana, the number of candidates is not so large, but since all are represented by a kana, it is difficult to find a correct answer candidate at a glance. In addition, there is an increased possibility that a pseudonym with a similar notation is mistaken and a wrong candidate is selected.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problem. When displaying candidates for a speech recognition result, the number of candidates is suppressed to an appropriate number, and a voice which is easy to find a correct answer candidate is displayed. An object of the present invention is to provide a recognition result display device.

[0006]

In order to achieve this object, a voice recognition result display device according to the present invention comprises a storage means for storing readings of kana-kanji mixed candidates in association with the kana-kanji mixed candidates, and a plurality of kana-kanji mixed candidates. And reading control means for controlling the display of kana-kanji mixed candidates having the same reading based on whether the kana-kanji candidate has already been displayed. .

[0007]

In the voice recognition result display device of the present invention having the above-described configuration, when displaying kana-kanji mixed candidates, the reading comparison means reads the candidates already displayed while referring to the reading information stored in the storage means. And the display control means controls the display of the candidate based on whether or not a kana-kanji mixed candidate having the same reading has already been displayed.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings.

First, the overall configuration will be described with reference to FIG. The speech recognition result receiving unit 10 is connected to a CPU (Central Processing Unit) 12, which is a display control means for controlling the display of the speech recognition result. The CPU 12 has a RAM (random access memory) as storage means for storing kana-kanji mixture candidates (hereinafter simply referred to as kana-kanji mixture candidates) as speech recognition results and reading, priority, etc. of the candidates. 14. ROM storing a program for comparing readings and controlling display
(Read only memory) 16 is connected. Also,
The CPU 12 is connected to a display device 18 for displaying kana-kanji mixed candidates as a speech recognition result.

When the speech recognition result receiving unit 10 receives the speech recognition result, the CPU 12 writes the kana-kanji mixture candidate as the speech recognition result and the reading thereof into the RAM 14 in association with each other. The kana-kanji mixed candidates stored in the RAM 14 are displayed on the display device 18 according to their priorities.
At this time, the CPU 12 compares the reading of the candidate to be displayed with the reading of the already displayed candidate according to the procedure stored in the ROM 16, and if a candidate having the same reading is already displayed, the candidate is displayed on the display device 18. Do not display.

Next, a procedure for displaying a speech recognition result will be described with reference to a flowchart of FIG. As an example,
After inputting "I bridged the river.", The case where the voice recognition result when saying "Hashiwa" is displayed will be described.
At this time, the kana-kanji mixed candidates and the readings of the voice recognition result stored in the RAM 14 are shown in FIG. Also,
At this time, candidates displayed on the display device 18 are shown in FIG.
(B). It is assumed that the candidates in FIG. 1A are already arranged in priority order.

Let i and j be counters for counting the number of candidates. First, i is set to 1 (S1). The i-th candidate is extracted from the RAM 14 (S2). In this example, the first candidate “foot” is extracted. Next, it is determined whether i is greater than 1 (S3). At first, since i = 1, the process jumps to S8 to display the i-th candidate on the display device 18.
To be displayed. In this example, the first candidate “foot”
Is displayed on the display device 18. Next, i is counted up by 1 (S9). In this example, i becomes 2. Then, if the i-th candidate still exists (S
10) Return to S2.

In step S2, the i-th candidate is extracted again. In this example, the second candidate "bridge" is extracted. Then, it is determined whether i is greater than 1 (S3). In this example, since i = 2, i is greater than 1 and S
Move to 4. j is set to 1 (S4), and the readings of the i-th candidate and the j-th candidate are compared in accordance with the procedure stored in the ROM 16 (S5). In this example, the reading of the second candidate “Hashiwa” is compared with the reading of the first candidate “foot”. Since "ashi" and "ashi" are not the same, the process jumps to S6.
j is incremented by 1 (S6), and it is determined whether i and j are the same (S7). In this example, j = 2,
i and j are the same. If they are the same, the reading is different from all the already displayed candidates, and the ith candidate is displayed on the display device 18. In this example, the second candidate “bridge” is displayed. Then, i is counted up (S9), and it is checked whether the i-th candidate exists (S10). In this example, i = 3, and the process returns to S2 because the third candidate “end” exists.

The third candidate "end" is taken out and the above operation is repeated. However, this time when reading is compared in S5, when j = 2, the i (= 3) th candidate is
The readings of the j (= 2) th candidate are both "Hashiha", and the readings are the same. Therefore, the process jumps from S5 to S9, i is counted up without displaying “end”, and the process proceeds to the fourth candidate “chopstick”. That is, if a candidate having the same reading has already been displayed, the candidate will not be displayed on the display device 18.

In this way, the candidates are displayed, and the processing is continued until there are finally no candidates for the speech recognition result in S10.

The embodiment of the present invention is not limited to the above example. For example, if a candidate having the same reading is already displayed, the candidate is not displayed but the same reading is performed. Modifications that do not deviate from the scope of the claims, such as displaying next to candidates having the same, are possible.

[0017]

As is apparent from the above description, in the speech recognition result display device of the present invention, the display of candidates having the same reading is controlled by whether or not the candidates are already displayed. Since the number is suppressed to about the same number as when displayed as a reading kana candidate, and is displayed not as a reading kana but as a kana-kanji mixed candidate, there is an effect that it is easy to find a correct candidate.

[Brief description of the drawings]

FIG. 1 is a diagram showing kana-kanji mixed candidates that are the result of speech recognition, their readings, and candidates displayed on a display device.

FIG. 2 is a block diagram showing an overall configuration.

FIG. 3 is a flowchart illustrating an operation of controlling display of candidates.

[Explanation of symbols]

 Reference Signs List 10 Voice recognition result receiving unit 12 CPU 14 RAM 16 ROM 18 Display device

Claims (1)

(57) [Claims] 1. A speech recognition result display device comprising a display unit for displaying kana-kanji mixture candidates of a speech recognition result, a storage unit for storing the reading of the kana-kanji mixture candidates in association with the kana-kanji mixture candidates, Reading comparison means for comparing the readings of kana-kanji mixed candidates to determine whether the readings are the same, and display control means for controlling the display of kana-kanji mixed candidates based on whether or not kana-kanji mixed candidates having the same reading are already displayed. A speech recognition result display device comprising: