JP2017212667A - Language information providing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a language information providing device capable of providing language information whose content is easy to grasp.SOLUTION: A device 10 comprises: a character number calculation part 13 which calculates the upper limit of the number of characters of language information to be output, on the basis of the output period of the language information; a language information generation part 15 which generates the language information to be output; and an original generation part 14 which, in accordance with the upper limit of the number of characters calculated by the character number calculation part 13, restricts the number of characters of the language information generated by the language information generation part 15.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a language information providing apparatus.

  For example, Patent Document 1 discloses a technique for reproducing subtitles of a translation sentence such as audio in a video together with the video.

JP 2009-16910 A

Taku Kudo, Yuji Matsumoto “Japanese Dependency Analysis by Chunking Stage Application” Transactions of Information Processing Society of Japan, Vol. 43 No. 6 June 2002, pp. 1834-1842

  It is known that the upper limit number of subtitles that can be displayed per unit time (hereinafter referred to as “the number of characters that can be displayed”) in consideration of the interpretation ability of a video viewer (user) is generally determined. For example, in the case of a movie, it is about 4 characters / second. The same can be said for scenes in which audio such as dubbing is synthesized with video as well as subtitles. In this case, the upper limit number of characters that can be output for dubbing audio per unit time corresponds to the number of displayable characters described above. obtain.

  When subtitles are displayed on a video such as a movie, the subtitles corresponding to each scene of the video must be displayed within the period of the scene. It is preferable that the number of subtitle characters per unit time displayed during the period is equal to or less than the number of displayable characters. However, when creating and displaying subtitles of translated sentences in real time (hereinafter sometimes simply referred to as “translated sentences”) in accordance with video playback, the number of translated characters exceeds the number of displayable characters. Can occur. Even when the dubbing voice is created and output, a situation may occur in which the number of characters of the dubbing voice exceeds the number of displayable characters. If a translated sentence exceeding the number of displayable characters is displayed or a dubbing voice having more characters than the number of displayable characters is output, it becomes difficult to grasp the contents of the language information represented by subtitles or dubbing voices. The same can be said for language information such as character information (transcript sentence) created from speech information by so-called transcription.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a language information providing apparatus capable of providing language information whose contents are easily grasped.

  A language information providing apparatus according to an aspect of the present invention includes: a calculation unit that calculates an upper limit number of characters of language information to be output based on an output time of language information; a creation unit that generates language information to be output; Limiting means for limiting the number of characters of the language information created by the creating means according to the upper limit number of characters calculated by the means.

  In the language information providing apparatus, the upper limit number of characters of the language information is calculated, and the number of characters of the language information to be created is limited according to the calculated upper limit number of characters. By limiting the number of characters of the language information in this way, it is possible to easily grasp the contents of the output language information.

  The calculation means calculates the display time of the caption in the video as the output time of the language information, calculates the upper limit number of characters of the displayed caption as the upper limit number of characters of the language information, and the creation means creates the translation sentence used as the caption as the language information Then, the restricting means may restrict the number of characters of the translation sentence created by the creating means. In this case, the upper limit number of characters of the displayed subtitle is calculated, and the number of characters of the translation to be created is limited according to the calculated upper limit number of characters. By limiting the number of characters in the translated sentence in this way, the translated sentence can be made easier to read.

  The restricting means may relax the restriction when a predetermined phrase or word is included in the language information. For example, if a phrase or word that can be grasped in a short period of time by the user becoming accustomed to reading or accustomed to listening is set as a predetermined phrase or word, such a phrase or word can be used as a language. Even if the above-mentioned restrictions are relaxed and the number of characters of the language information is increased, it is possible to prevent the language information from becoming difficult to grasp.

  The restricting means limits the number of characters in the language information by preventing the number of characters in the language information from exceeding the upper limit number of characters when the predetermined phrase or word is not included in the language information. When a phrase or a word is included in the language information, the restriction may be relaxed by allowing the number of characters in the language information to exceed the upper limit number of characters. For example, in this way, the number of characters in the language information can be restricted and the restriction can be relaxed.

  The restricting means creates an original text that is the basis of the language information, the creating means creates language information output from the original text created by the restricting means, and the restricting means shortens the original text, thereby May be performed. For example, the number of characters in the language information can be limited by shortening the original text in this way.

  ADVANTAGE OF THE INVENTION According to this invention, the linguistic information which can grasp | ascertain the content easily can be provided.

It is a figure which shows schematic structure of the apparatus which concerns on embodiment. It is a figure which shows the example of the hardware constitutions of an apparatus. It is a flowchart which shows an example of the process performed in an apparatus. It is a flowchart which shows an example of the process performed in an apparatus. It is a figure which shows the example of the translation sentence by which the number of characters was displayed displayed as a caption. It is a figure which shows schematic structure of the apparatus which concerns on a modification.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant descriptions are omitted.

  FIG. 1 is a diagram illustrating a schematic configuration of a language information providing apparatus (hereinafter also simply referred to as “apparatus”) according to an embodiment. The apparatus 10 shown in FIG. 1 may be, for example, an apparatus (translation apparatus) that outputs, as language information, subtitles of a translation sentence of an audio in a video including a moving image and audio together with the video. In that case, in the device 10, a translation is created and output together with the video when the video is played back. Audio (dubbing audio) corresponding to the translated sentence instead of subtitles may be created as language information and output together with the video. Even without translation, a transcript sentence transcribed for a hearing impaired person in a lecture or the like may be created and output as language information. In the case of a transcript statement, the video does not have to be output. The device 10 can be realized in various electronic devices such as a television, an electric bulletin board, a digital signage, a PC, a smartphone, and a robot capable of communicating with a user.

  As shown in FIG. 1, the apparatus 10 includes an input unit 11, a speech recognition processing unit 12, a character count calculation unit 13, a source text creation unit 14, a language information creation unit 15, an automatic summarization unit 16, and a storage Part 17 and output part 18.

  The input unit 11 is a part for inputting audio and video. For example, audio and video data (audio data and video data) are acquired by a camera, a microphone, or the like that can constitute the input unit 11. The audio data and video data may be acquired by the input unit 11 from a content server that distributes video via a communication network such as the Internet. In addition, the input unit 11 receives a user operation for designating which information of subtitles, dubbed voices, and transcript sentences should be output as language information. For example, an instruction to output subtitles or voice-over audio is input by the input unit 11. Alternatively, the input unit 11 inputs an instruction to execute processing (transcription processing) for outputting a transcript sentence.

  The voice recognition processing unit 12 is a part that recognizes the voice input by the input unit 11. Speech recognition is to specify the meaning and contents of the entire speech input by the input unit 11 or each word included in the speech. Various known speech recognition engines capable of performing speech recognition processing may be used for speech recognition. In that case, a speech recognition engine having different types of acoustic models and language models may be used depending on the video attributes.

  The number-of-characters calculation unit 13 is a part (calculation unit) that calculates the upper limit number of characters of the language information to be output when language information such as the above-described subtitles, dubbed voices, and transcript sentences is output. For example, the number-of-characters calculation unit 13 calculates the maximum allowable number of characters of subtitles displayed in video or output dubbed voice as the upper limit number of characters. The maximum allowable number of characters for subtitles or dubbed audio is the upper limit of the number of characters in the translated text to be displayed in one scene of the video, taking into account the reading ability of the video viewer (in this example, the user of the device 10). As one scene of the video, a scene where a certain person in the video describes one line is assumed. In addition, the character number calculation unit 13 calculates the maximum allowable number of characters of a transcript sentence displayed at a lecture or the like as the upper limit number of characters. The maximum allowable number of characters in a transcript sentence is the upper limit of the number of characters in the transcript sentence that should be displayed in one scene such as a lecture, taking into account the viewer's ability to interpret the transcript sentence. As a scene such as a lecture, a scene where a person who is giving a speech describes a single line is assumed.

  An example of calculating the maximum allowable number of characters will be described. When the maximum allowable number of characters is the upper limit number of characters of subtitles or dubbed audio, for example, the character number calculation unit 13 multiplies the subtitle display time or dubbed audio output time in one scene of the video by the above-described number of displayable characters. Thus, the maximum allowable number of characters is calculated. In order to calculate the maximum allowable number of characters, the number-of-characters calculation unit 13 acquires the subtitle display time or dubbing sound output time in the video input by the input unit 11. The acquisition of the display time or the like may be acquired, for example, by preparing data specifying the display time or the like in advance and inputting the data together with the audio data and the video data by the input unit 11. When there is no data specifying the display time or the like, the character number calculation unit 13 may measure the display time or the like. Measurement of display time and the like is performed by analyzing video data. For example, the number-of-characters calculation unit 13 measures the time during which speech is uttered as sound in one scene of the video, and sets the measured time as the caption display time in the video. When the maximum allowable number of characters is the upper limit number of characters in the transcript sentence, for example, the character number calculation unit 13 multiplies the display time of the transcript sentence in one scene such as a lecture by the above-described number of characters that can be displayed. Calculate the number of characters. The display time of the transcript sentence may be acquired by the character number calculation unit 13 in the same manner as the display time of the caption described above.

  The original sentence creation unit 14 is a part that creates an original sentence that is the basis of language information created by the language information creation unit 15 described later. The original sentence created first by the original sentence creation unit 14 is a sentence corresponding to the voice input by the input unit 11. For example, when audio data input together with video by the input unit 11 is given as character data (text data), the original text is created by using the character data as it is. On the other hand, if there is no such character data, the original text creation unit 14 may create text data corresponding to the voice input by the input unit 11 and recognized by the voice recognition processing unit 12 as the original text. Good.

  The language information creating unit 15 is a part (creating unit) that creates language information to be output from the original text created by the original text creating unit 14. For example, when the language information is the above-described subtitle or dubbing voice, the language information creation unit 15 creates a translation of the original text created by the original text creation unit 14. Various known machine translation techniques may be used for translation. The translated sentence created by the language information creating unit 15 can be used as a caption displayed on the video. In addition, the language information creating unit 15 creates sound corresponding to the translated sentence (synthesized sound output together with the video). The synthesized speech may be automatically created from the translated sentence, or may be created by reading out the translated sentence by a person (reader). In the case of automatically generating synthesized speech, various known methods for converting character data into speech data may be used. When the reader creates the synthesized speech, the translated sentence created by the language information creating unit 15 is displayed by, for example, the output unit 18 described later, and the reader reads the displayed translated sentence. The sound corresponding to the translated sentence created by the language information creating unit 15 can be used as the dubbed sound output together with the video. When the language information is the above-described transcript sentence, the language information creating unit 15 creates a transcript sentence by using the original sentence created by the original sentence creating unit 14 as it is.

  The automatic summarization unit 16 is a part that creates a summary sentence of the original text created by the original text creation unit 14. The summary sentence can be created by shortening the original sentence step by step. A method for creating such a summary sentence is not particularly limited, but the method described in Non-Patent Document 1 can be used as an example.

  The above-described original text creation unit 14, language information creation unit 15, and automatic summarization unit 16 are also portions (limitation means) that limit the number of characters of language information according to the maximum allowable number of characters calculated by the number of characters calculation unit 13. An example of the restriction method will be described. First, the language information creation unit 15 translates the original text created by the original text creation unit 14. The source text creation unit 14 counts the number of characters of the language information created by the language information creation unit 15 and determines whether or not the counted number of characters of the language information exceeds the maximum allowable number of characters calculated by the number of characters calculation unit 13. . When the number of characters in the translated sentence exceeds the maximum allowable number of characters, the original sentence creating unit 14 instructs the automatic summarizing unit 16 to summarize the original sentence. The original sentence creation unit 14 creates a summary sentence obtained by summarization by the automatic summarization part 16 as a new original sentence. Since the newly created original text corresponds to the summary text of the original text created first, it is shorter than the original text created first. The language information creation unit 15 creates language information from the original text newly created by the original text creation unit 14. Thus, the number of characters of the language information obtained is smaller than the number of characters of the language information created first. By reducing the number of characters in the language information in this way, the number of characters in the language information can be limited. As described above, the summarization by the automatic summarization unit 16 can be performed in stages, so that the linguistic information can be shortened in stages by repeatedly executing the above series of processes. Then, by repeating such a series of processes until the number of characters in the language information becomes equal to or less than the maximum allowable number of characters, the number of characters in the language information is limited to the maximum allowable number of characters.

  Furthermore, in the present embodiment, when a predetermined phrase or word is included in the language information created by the language information creation unit 15, the limitation on the number of characters in the language information is relaxed. Predetermined phrases and words are phrases and words that allow the user to grasp the contents in a short time compared to other phrases and words when the user is accustomed to reading or listening. . As an example of the phrase, there is a standard phrase “good morning” used daily. Examples of words include the proper noun “New Zealand” that is used on a daily basis.

  The original text creating unit 14 determines whether or not a predetermined phrase or word is included in the language information. When the predetermined phrase or word is not included in the language information, the original text creating unit 14 cooperates with the language information creating unit 15 and the automatic summarizing unit 16 as described above, and the number of characters of the language information. Limit the number of characters in the language information so that does not exceed the maximum allowable number of characters. On the other hand, when a predetermined phrase or word is included in the linguistic information, the original text creation unit 14 relaxes the restriction on the number of characters in the language information by allowing the number of characters in the translated sentence to exceed the maximum allowable number of characters. To do. For example, the original text creation unit 14 calculates the number of characters obtained by subtracting the total number of characters of a predetermined phrase or word included in the language information from the number of characters of the language information as the number of characters of the new language information. By comparing the calculated number of characters of the new language information with the maximum allowable number of characters, the source text creation unit 14 cooperates with the language information creation unit 15 and the automatic summarization unit 16 as described above, and Limit the number of characters in language information so that the number of characters does not exceed the maximum number of characters.

  The storage unit 17 is a part that stores information necessary for various processes executed in the apparatus 10. In particular, the storage unit 17 stores the above-described number of displayable characters and the above-described predetermined phrases and words.

  The output unit 18 is a part that outputs the language information created by the language information creation unit 15. When the language information is a translated sentence of the original sentence, the output unit 18 sets the translated sentence of the original sentence as a caption and outputs it with the video input by the input unit 11. When the language information is dubbed sound, the output unit 18 outputs the dubbed sound together with the video. When the language information is a transcript sentence, the output unit 18 displays the original sentence.

  The language information creating unit 15 is a part (creating unit) that creates language information to be output from the original text created by the original text creating unit 14. For example, when the language information is the above-described subtitle or dubbing voice, the language information creation unit 15 creates a translation of the original text created by the original text creation unit 14. Various known machine translation techniques may be used for translation. The translated sentence created by the language information creating unit 15 can be used as subtitles displayed on the video, and the audio corresponding to the translated text can be used as dubbed audio output together with the video. When the language information is the above-described transcript sentence, the language information creating unit 15 creates a transcript sentence by using the original sentence created by the original sentence creating unit 14 as it is. When the language information is a transcript sentence, it is only necessary to output at least the transcript sentence, and the output of the video is not essential. An example of subtitles and videos output by the output unit 18 will be described later with reference to FIG.

  FIG. 2 is a diagram illustrating an example of a hardware configuration of the apparatus. As shown in FIG. 2, the device 10 is physically composed of one or more CPUs (Central Processing Units) 101, a RAM (Random Access Memory) 102 and a ROM (Read Only Memory) 103, which are main storage devices, It can be configured as a computer including a communication module 104 that is a data transmission / reception device, an auxiliary storage device 105 such as a semiconductor memory, an input device 106 that receives input of a user operation, an output device 107 such as a display, and the like. The functions of the apparatus 10 described above with reference to FIG. 1 communicate with each other under the control of the CPU 101 by, for example, loading one or more predetermined computer software on the hardware such as the CPU 101 and the RAM 102. This can be realized by operating the module 104, the input device 106, the output device 107, and the like, and reading and writing data in the RAM 102 and the auxiliary storage device 105.

  FIG. 3 is a flowchart illustrating an example of processing executed in the apparatus. The process of this flowchart is executed, for example, in one scene of a video being reproduced or one scene in a lecture. When the process is executed in one scene of the video, a process involving translation is executed (steps S4 to S9 described later). When the process is executed in one scene such as a lecture, a transcription process without translation is executed (steps S10 to S13 described later). Which process is branched to depends on the content of the instruction by the user operation received by the input unit 11 as described above. It is assumed that the user operation is accepted by the input unit 11 prior to the start of the flowchart of FIG.

  In step S1, the apparatus 10 acquires the output time of language information. Specifically, as described above, based on the voice or the like input by the input unit 11, the character number calculation unit 13 determines the subtitle display time, dubbing voice output time, or transcript sentence display time. , Get as language information output time.

  In step S2, the apparatus 10 calculates the maximum allowable number of characters from the output time. Specifically, the number-of-characters calculation unit 13 stores the subtitle display time (seconds), dubbing voice output time (seconds) or transcript sentence output time (seconds) calculated in the previous step S1 in the storage unit 17. The maximum allowable number of characters is calculated by multiplying the stored number of displayable characters (for example, 4 characters / second).

  In step S3, the apparatus 10 determines whether translation or transcription should be performed. For example, when an instruction to output subtitles or dubbed audio is input by the input unit 11 by a user operation of the device 10, it is determined that translation should be performed because translation processing is necessary. On the other hand, when an instruction to execute the transcription process is input by the user operation through the input unit 11, it is determined that the transcription should be performed. If it is determined that translation is to be performed, the apparatus 10 proceeds to steps S4 to S9. If it is determined that transcription should be performed, the apparatus 10 proceeds to steps S10 to S13 (FIG. 4 described later).

  In step S4, the apparatus 10 performs machine translation. Specifically, the automatic summarization unit 16 creates a translation of the original text created by the original text creation unit 14 as described above. The process of step S4 is repeatedly executed when the process of step S8 described later is executed, but the original text that is the basis of translation in the first process of step S4 is as described above. The original text created first by the original text creation unit 14. In the process of step S4 after the second time, the original text shortened by the process of step S8 described later is created as a new original text.

  In step S5, the apparatus 10 determines whether or not an exceptional phrase or word is included in the translated sentence. Specifically, the original sentence creation unit 14 compares the translation sentence created in the previous step S4 with the phrase or word stored in the storage unit 17, and whether the phrase or word is included in the storage unit 17 or not. Judge whether or not. If the translated sentence includes an exceptional phrase or word (step S5: YES), the apparatus 10 advances the process to step S6. When that is not right (step S5: NO), the apparatus 10 skips step S6 and advances a process to step S7.

  In step S6, the apparatus 10 subtracts the exceptional phrase or word from the number of translated characters. Specifically, the original sentence creating unit 14 newly adds the number obtained by subtracting the number of characters of the phrase or word determined to be included in the translated sentence in the previous step S5 from the number of characters of the translated sentence created in the previous step S4. Calculated as the number of characters in a translated sentence. When the number of characters in the new translated sentence is calculated in this way, even if the number of characters in the translated sentence before the new calculation exceeds the maximum allowable number of characters, a YES determination can be made in step S7 described later, and a step described later In S8, the translated sentence having the number of characters exceeding the maximum allowable number of characters can be output together with the video as subtitles or the sound corresponding to the translated sentence as dubbed sound.

  In step S7, the apparatus 10 determines whether or not the number of translated characters is equal to or less than the maximum allowable number of characters. Specifically, the number of characters of the translated sentence created in the previous step S4 or the number of characters of the new translated sentence calculated in the previous step S6 is determined by the original sentence creating unit 14 as the maximum allowable number of characters calculated in the previous step S2. It is determined whether or not: If the number of translated characters is less than or equal to the maximum allowable number of characters (step S7: YES), the apparatus 10 advances the process to step S9. When that is not right (step S7: NO), the apparatus 10 advances a process to step S8.

  In step S8, the apparatus 10 shortens the sentence step by step. Specifically, the automatic summarization unit 16 summarizes the original text created in the previous step S4. The original sentence creation unit 14 creates a summary sentence obtained by summarization by the automatic summarization part 16 as a new original sentence. After the process of step S8 is completed, the apparatus 10 returns the process to step S4 again. As described above, a translation of the original sentence is created in step S4. Further, as shown in FIG. 3, the processes in steps S4 to S8 can be repeatedly executed. By repeatedly executing these processes, the translation sentence is shortened step by step along with the original sentence.

  In step S9, the apparatus 10 performs subtitle display / speech synthesis or transcript statement display. Specifically, the output unit 18 displays the translated sentence created in the previous step S4 as subtitles together with the video, or outputs the audio corresponding to the translated sentence as dubbed audio together with the video. Alternatively, the original sentence created by the original sentence creating unit 14 or the original sentence newly created in step S13 described later is output as a transcript sentence. After the process of step S9 is completed, the apparatus 10 ends the process of the flowchart.

  Next, processing that is executed when it is determined in step S3 that transcription should be performed and the processing proceeds to step S10 shown in FIG. 4 will be described.

  In step S10, the apparatus 10 determines whether or not the sentence includes an exceptional phrase or a word. Specifically, the original sentence creation unit 14 compares the original sentence that can be a transcript sentence created as described above with a phrase or word stored in the storage unit 17, and the phrase or word is stored in the storage unit 17. It is judged whether it is included in. If the original sentence includes an exception phrase or word (step S10: YES), the apparatus 10 proceeds to step S11. When that is not right (step S10: NO), the apparatus 10 skips step S11 and advances a process to step S12.

  In step S11, the apparatus 10 subtracts the exceptional phrase or word from the number of characters. Specifically, the original text creation unit 14 calculates the number of characters of the original text by subtracting the number of characters of the phrase or word determined to be included in the original text in the previous step S10 as the number of characters of the new original text. When the number of characters of the new original text is calculated in this way, even if the number of characters of the original text before the new calculation exceeds the maximum allowable number of characters, a YES determination can be made in step S12 described later, and the steps executed later In S9, an original sentence having a number of characters exceeding the maximum allowable number of characters can be output as a transcript sentence.

  In step S12, the apparatus 10 determines whether the number of characters is equal to or less than the maximum allowable number of characters. Specifically, the original text creation unit 14 determines whether the number of characters in the original text or the number of characters in the new original text calculated in the previous step S11 is less than or equal to the maximum allowable number of characters calculated in the previous step S2. If the number of characters is less than or equal to the maximum allowable number of characters (step S12: YES), the apparatus 10 advances the process to step S9 (FIG. 3) described above. When that is not right (step S12: NO), the apparatus 10 advances a process to step S13.

  In step S13, the apparatus 10 shortens the sentence step by step. Since this process is the same as the process of the previous step S8, description thereof is omitted here. After the process of step S13 is completed, the apparatus 10 returns the process to step S10 again.

  According to the apparatus 10 described above, the upper limit number of characters of language information is calculated (step S2), and the number of characters of language information to be created is limited according to the calculated upper limit number of characters (steps S4 to S8, S10). S13). By limiting the number of characters of the language information in this way, it is possible to easily grasp the contents of the output language information (step S8).

  For example, the upper limit number of subtitle characters displayed in the video is calculated (step S2), and the number of characters of the translation to be created is limited according to the calculated upper limit number of characters (steps S3 to S8). By limiting the number of characters of the translated text displayed in this way, when the translated text with the limited number of characters is displayed as a subtitle along with the video (step S9), the translated text becomes easy to read.

  Further, when a predetermined phrase or word is included in the language information (step S5: YES, step S10: YES), the above-described restriction is relaxed (steps S6 and S11). Here, a phrase or a word that can be grasped in a short time by the user becoming accustomed to reading or accustomed to listening is set as a predetermined phrase. Therefore, even if such a phrase or word is included in the language information and the above-mentioned restriction is relaxed and the number of characters of the language information is increased (step S8), it is suppressed that the language information becomes difficult to grasp. be able to.

  If a predetermined phrase or word is not included in the linguistic information (step S5: NO, step S10: NO), the linguistic information can be stored by making the number of characters in the linguistic information not exceed the upper limit number of characters. The number of characters is limited (step S7: NO, step S9, step S12: NO, step S13). On the other hand, when a predetermined phrase or word is included in the translated sentence (step S5: YES, step S10: YES), the above-mentioned restriction is allowed by allowing the number of characters in the translated sentence to exceed the upper limit number of characters. Relaxed (step S6, step S7: YES, step S11, step S12: YES). In this way, the number of characters in the language information can be restricted and the restriction can be relaxed.

  Moreover, the above-mentioned restriction | limiting is performed by shortening the original text used as the foundation of language information (step S8, S13). By shortening the original text in this way, the number of characters in the language information can be limited.

  FIG. 5 is a diagram illustrating an example of a translated sentence with a limited number of characters displayed as subtitles. The translated sentence displayed on the screen D1 on the left side of FIG. 5 is a comparative example, and the translated sentence displayed on the screen D2 on the right side of FIG. 5 is a caption whose number of characters is limited by the device 10. . These screens D1 and D2 are screens that can be output by the output unit 18, and are specifically screens displayed on the output device 107 such as a display as described above with reference to FIG.

  First, a screen D1 as a comparative example will be described. On the screen D1, a translated sentence whose number of characters is not limited is displayed. In this example, in the scene where the person in the video says the original sentence, “I was really happy to come over last weekend. It is displayed with the video. On the other hand, a translated sentence with a limited number of characters is displayed on the screen D2. Comparing screen D2 with screen D1, screen D2 displays the translated text of the original text summarized by deleting the words “Come to me” and “Mata” from the words of the person in the video. Yes. As a result, the number of characters of the translated text displayed on the screen D2 is limited, and it is easy for the user to read. By limiting the number of characters, the translated text displayed on two lines on the screen D1 is also displayed on one line on the screen D2. As a result, the user can comfortably read the subtitles.

  Although one embodiment of the present invention has been described above, the present invention is not limited to the above embodiment.

  For example, a part of the processing executed in the device may be executed in the external server. FIG. 6 is a diagram showing a schematic configuration of an apparatus 10A according to such a modification. Device 10A is different from device 10 (FIG. 1) in that it does not include speech recognition processing unit 12 and language information creation unit 15, but includes communication unit 19. The communication unit 19 is a part for communicating with the server 20. The server 20 includes a voice recognition processing unit 21, a language information creation unit 22, and a communication unit 23. Since the functions of the speech recognition processing unit 21 and the language information creation unit 22 are the same as those of the speech recognition processing unit 12 and the language information creation unit 15 described above, the description thereof is omitted here. The communication unit 23 is a part that communicates with the communication unit 19 of the apparatus 10A. The apparatus 10 </ b> A shown in FIG. 6 can use the functions of the speech recognition processing unit 21 and the language information creation unit 22 of the server 20 by communicating with the server 20 using the communication unit 19. Even with such a configuration, the apparatus 10A can execute the same processing as that of the apparatus 10 described above. In this case, the processing load on the apparatus 10A can be reduced by the amount of processing executed by the server 20 for speech recognition and language information creation.

  Moreover, in the said embodiment, 4 characters / second was mentioned as an example and demonstrated as the number of displayable characters. However, the number of displayable characters is not limited to this example. For example, a different number of displayable characters may be employed depending on the type of video. As video types, various videos such as movie videos, lecture videos, and conference videos are assumed. For example, in the case of a movie image, 3 to 4 characters / second may be adopted as the number of displayable characters. When the number of displayable characters is properly used according to the type of video, the storage unit 17 may store in advance a data table in which the type of video and the number of displayable characters are described in association with each other. Then, for example, the number-of-characters calculation unit 13 may acquire the number of displayable characters corresponding to the type of video from the storage unit 17 and calculate the maximum allowable number of translated characters using the acquired number of displayable characters. When the translated sentence is created and displayed in a language different from Japanese, a different number of displayable characters may be employed depending on the language. In this case as well, the number-of-characters calculation unit 13 and the storage unit 17 may be customized as in the case where the number of characters that can be displayed is properly used according to the type of video described above. In addition, the number of displayable characters may be properly used for each user of the device 10 or for each content.

  Moreover, in the said embodiment, when the predetermined phrase or word was contained in language information, the example which subtracts the number of characters of these phrases or a word from the number of characters of language information was demonstrated. In this case, instead of simply subtracting the number of characters in the phrase or word from the number of characters in the language information, correct the number of characters according to the type of phrase or word (for example, by multiplying by a correction factor) The number of characters after may be subtracted from the number of characters in the language information. In that case, it is better to increase the number of corrected characters (for example, increase the correction coefficient) as the phrase or word becomes more familiar to the user or familiar to the user. When the degree of subtraction from the number of characters in the language information is changed according to the type of phrase or word, the storage unit 17 stores in advance a data table in which the type of phrase or word is described in association with, for example, a correction coefficient. It is good to keep. Then, the number-of-characters calculation unit 13 acquires a correction coefficient corresponding to the type of the phrase or word from the storage unit 17, and subtracts the number of characters obtained by multiplying the acquired correction coefficient by the number of characters of the phrase or word from the number of characters of the language information. .

  Moreover, in the said embodiment, the number-of-characters calculation part 13 measures the time when the speech is emitted as audio | voice in one scene of an image | video as needed, and makes the measured time the display time of a subtitle in an image | video as needed. Explained. However, the display time of the subtitles may be set so that the display time of the corresponding subtitles is longer than the time when the speech is being emitted. Conversely, the display time of the subtitles may be set so that the display time of the corresponding subtitles is shorter than the time when the speech is being emitted.

  Moreover, in the said embodiment, the original sentence preparation part 14 performs the judgment of whether the number of characters of language information exceeds the maximum number of allowable characters, or whether a predetermined phrase or word is included in language information, etc. Explained. However, these determinations may be made by elements other than the original text creation unit 14, such as the character count calculation unit 13, the language information creation unit 15, and the automatic summarization unit 16.

  DESCRIPTION OF SYMBOLS 10, 10A ... Apparatus, 11 ... Input part, 12, 21 ... Speech recognition processing part, 13 ... Character number calculation part, 14 ... Original text creation part, 15, 22 ... Language information creation part, 16 ... Automatic summary part, 17 ... Memory | storage Part, 18 ... output part, 19, 23 ... communication part, 20 ... server.

Claims (5)

Calculation means for calculating the upper limit number of characters of the language information to be output based on the output time of the language information;
Creating means for creating the output language information;
Limiting means for limiting the number of characters of the language information created by the creating means according to the upper limit number of characters calculated by the calculating means;
Comprising
Language information providing device. The calculation means calculates the display time of the caption in the video as the output time of the language information, calculates the upper limit number of characters of the displayed caption as the upper limit number of characters of the language information,
The creating means creates a translation used as the subtitle as the language information,
The restricting means restricts the number of characters of the translation sentence created by the creating means;
The language information providing apparatus according to claim 1.   The language information providing apparatus according to claim 1, wherein the restriction unit relaxes the restriction when a predetermined phrase or word is included in the language information. The limiting means is
When the predetermined phrase or word is not included in the language information, by limiting the number of characters of the language information by preventing the number of characters of the language information from exceeding the upper limit number of characters,
When the predetermined phrase or word is included in the language information, the limit is relaxed by allowing the number of characters in the language information to exceed the upper limit number of characters.
The language information providing apparatus according to claim 3. The restriction means creates an original text that is the basis of the language information,
The creation means creates the output language information from the original text created by the restriction means,
The restriction means performs the restriction by shortening the original text.
The language information providing apparatus according to any one of claims 1 to 4.

Images