JP2015125232A - Voice recognition error correction device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a voice recognition error correction device capable of reducing generation of an automatic correction error in a conventional block collation system.SOLUTION: A voice recognition error correction device 100 comprises: WFST storage means 110 in which a correspondence manuscript set (WFST) constructed by preliminarily reading a manuscript text set 200 is stored; node data update means 120 for calculating and updating scores in a state of being able to be transited on a network of the WFST as node data for every time for accepting input of words of a recognition word string of a voice recognition device 220; node data storage means 130; manuscript search means 140 for sequentially determining hypotheses partially approximating the final best hypothesis as error correction results while tracing back on the network of the WFST based on the node data which is stored at that point of time whenever a processing start condition is satisfied; and manuscript output means 150 for sequentially outputting corresponding manuscripts determined as the error correction results.

Description

  The present invention relates to a speech recognition technique, and more particularly to a speech recognition error correction device that automatically corrects recognition errors.

  For example, when subtitles are generated by recognizing broadcast audio in a TV broadcast news program, the operator visually checks the speech recognition result presented on the touch panel monitor, for example, and corrects if an error is found. Doing work. Patent Document 1 discloses a technique for automatically correcting a speech recognition error by reducing the load on the operator. For example, the recognition result of the utterance voice of the news program is the recognition result of the voice uttered based on the manuscript text as the utterance information source, and thus such automatic correction can be performed.

  However, although the utterance content that is the target of speech recognition is based on the manuscript, some words may be rephrased, some phrases may be dropped, or new information may be added. On the other hand, since the speech recognition apparatus uses an adaptation made using this manuscript text, the same part as the manuscript in the utterance word string can be recognized with high accuracy, but the portion that does not match the manuscript has a recognition accuracy. descend. It also includes recognition errors caused by noise, unclear utterances, misrepresentations, and stagnation.

The technique described in Patent Document 1 uses a word chain block having a length N (number of words N) in a section of a document to be determined as an output candidate in the document from the speech recognition result. The section of the manuscript to be determined as an output candidate is a section in which the word string of the speech recognition result is compared with the word string in the manuscript to obtain the matching rate of both word strings. In the technique described in Patent Document 1, it is assumed that the word string (word count N _h ) of the speech recognition result and the word string (word count N _r ) in the manuscript both have the same word count N. For the word sequence (word count N = N _h ) of the speech recognition result to be searched, the word sequence of the manuscript that has the smallest mismatch rate is searched, and a part thereof is output as the correction result. Hereinafter, this conventional method is referred to as a block matching method.

For simplicity, the block matching method will be described with reference to FIG. 8 with N = N _h = 6. One of the manuscript texts to be spoken is the original manuscript 600 “The Japan Meteorological Agency last month on the Pacific side of West Japan and East Japan…”. In this case, the collation block B1 is set to include 6 words from “west” of the original document 600, and the collation document 701 is a chain block of 6 words from “west” to “no”. The collation block B2 is set to be shifted to the right by one word from the collation block B1, and the collation document 702 is a chain block of six words from “Japan” to “Pacific”. Similarly, the collation block B3 is set so as to be shifted by one word from the collation block B2, and the collation document 703 is a chain block of six words from “to” to “side”. Further, the collation block B4 is set so as to be shifted by one word from the collation block B3, and the collation document 704 is a chain block of six words from “East” to “In”. Further, the collation block B5 is set by shifting one word from the collation block B4, and the collation document 705 is a chain block of six words from “Japan” to “Last month”.

  Here, for example, if the recognition result 801 read out from the original manuscript 600 and recognized by speech is “Last month on the Atlantic side of Japan”, the third recognition word “Atlantic” input is an error of “Pacific”. It is. When the verification blocks B1 to B5 are used, the verification document 705 has the highest matching rate with the recognition result 801. Therefore, in this case, the corrected output 901 is “the last month on the Pacific side of Japan”, which is the same text data as the collation manuscript 705. This corrects the error.

JP 2012-128188 A

  The block matching method, which is the prior art, can perform automatic correction, but has an automatic correction error due to an unknown block boundary of a word string as a recognition result, and there is room for improvement. In the previous example, in the original document 600, “Meteorological Agency” is arranged after the last word “last month” in the confirmed word string, so the word string input next to the recognition result 801 is collated. When block B6 is set to include 6 words from “Meteorological Agency” and collation manuscript 706 is a six-word chain block from “Meteorological Agency”, the matching rate between the next recognition result and the collation manuscript is the highest. It is thought to be higher. However, it was not known from the beginning that there was a block boundary between the word “Last Month” and the word “Meteorological Agency”.

For example, if a phrase or the like is dropped in part of the utterance content, the number of words N _{r in} the section of the manuscript to be determined as an output candidate is based on the number of words N = N _{h in} the word string of the speech recognition result as a reference. Will also grow. Here, if the number of words shed clauses and [delta] _d, a relationship of _{N r = N h + δ d} . Therefore, from the viewpoint of the reference word number N = N _h , the relationship is N = N _h <N _r . Would otherwise, in the original section of the longer N _r than the number of words N of the reference, but must find suggestions, such suitable N _r can not be set are unknown.

As a specific example, it is assumed that the recognition result 802 shown in FIG. 9 is obtained from the original document 600 shown in FIG. The recognition result 802 is “the Japan Meteorological Agency last month on the East Japan side”. Focusing on the block consisting of the 1st to 6th input words “East ... Last month”, the drop of “Pacific” occurs in the 4th compared to the original manuscript 600, and the recognition word “side” is input. That is, δ _d = 1. In this case, when searching for a collation document whose section length is 6 shown in FIG. 8 in order to obtain a document to be output, the collation document 704 has the highest matching rate with the recognition result 802 in FIG. It is. As a result, the missing “Pacific” is restored in the section of length 6 in the current collation. However, the following problem occurs in the collation block to be collated next. Corresponding document 706 has the highest matching rate corresponding to the block consisting of the seventh and subsequent input words “Meteorological Agency is ...” of recognition result 802. That is, when viewed from the corrected output 902 obtained by connecting the previous collation and the current collation, there is a problem that “last month” that is a recognition word at the block boundary is missing from the output.

When the content of the utterance includes repetition of addition of information or segmentation due to stagnation, the number of words N _{r in} the section of the manuscript to be determined as an output candidate is the number N of words in the word string of the speech recognition result as a reference. = it becomes smaller than N _h. Here, if the number of words in the added information or the number of words repeatedly spoken due to stagnation is δ _i , the relationship is N _r = N _h −δ _i . Therefore, in view of the number of words N = N _h as a reference, a relationship of N = N _h> N _r. Would otherwise, in the original section of the shorter N _r than the number of words N of the reference, but must find suggestions, such suitable N _r can not be set are unknown.

As a specific example, a case is assumed where the recognition result 803 shown in FIG. 10 is obtained from the original document 600 shown in FIG. The recognition result 803 is 6 words “To Taihei Yoga in Japan”. Compared with the original manuscript 600, the segment breaks (segments) are different, word errors have occurred, and only five words of the original manuscript are recognized. That is, δ _i = 1. In this case, when searching for a collation manuscript in which the length of the section shown in FIG. 8 is 6 in order to obtain a manuscript to be output, the matching score with the recognition result 803 in FIG. The two collation originals 705 and 704 are provided.

  For example, when the collation original 1 (collation original 705) is adopted, the correction output 1 by the current collation becomes the correction output 901 of the same text as the collation original 705. However, when “last month” is input as a recognition word following the recognition result 803, the score of the collation document is highest when the head of the collation block to be collated next time is “last month”, so the corrected output 1 The output immediately after (the output of the next collation) is the corrected output 903. In other words, when the corrected output obtained by concatenating the previous collation and the current collation is viewed, there is a problem that “last month” that is a recognition word at the block boundary is output twice.

  Similarly, when the collation document 2 (collation document 704) is employed, for example, the correction output 2 by the current collation is the same text correction output 904 as the collation document 704. However, if “... West Japan and East” is input as the recognition word before the input of the recognition result 803, the collation manuscript score when the last collation block collated immediately before is “east” is the highest. Therefore, the output immediately before the corrected output 2 (the output of the previous collation) becomes the corrected output 905. In other words, when the corrected output obtained by concatenating the previous collation and the current collation is viewed, there is a problem that “east”, which is a recognition word at the block boundary, is output twice.

  Furthermore, such inconsistency that occurs in the document section corresponding to the word sequence of the speech recognition result used as a reference is caused not only by the utterance content but also by a recognition error of the speech recognition apparatus. For example, an error such as recognizing a plurality of words as one word causes a discrepancy similar to that of a missing phrase. On the other hand, an error that recognizes one word as a plurality of words causes a mismatch similar to the addition of information. In particular, when there is a misrecognized word at the block boundary, the corresponding manuscript section becomes inappropriate and often cannot be corrected correctly.

  In the prior art, as a result of collating these inappropriate manuscript sections with the word string of the speech recognition result, an automatic operation such as missing a word or outputting the same word twice at a block boundary. A correction error occurs. Such an automatic correction error is the same even if the unit of the section in which the speech recognition result corresponds to the manuscript is set as a sentence or other unit as long as the boundary (sentence boundary) of the word hypothesis sequence based on the speech recognition result is unknown. Arise.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a speech recognition error correction apparatus that can reduce the occurrence of automatic correction errors in a conventional block matching method.

  In order to solve the above problems, a speech recognition error correction apparatus according to the present invention receives a recognition word string output from a speech recognition apparatus that recognizes an uttered speech read out from a document included in a document text set as input, and stores it in advance. A speech recognition error correction apparatus for correcting an error included in the recognized word string by estimating a word string of the corresponding document from the corresponding document set, comprising: a corresponding document set storage unit; a node data update unit; A node data storage unit, a document search unit, and a document output unit.

  According to such a configuration, the speech recognition error correction apparatus represents the state transition between the nodes representing the state of the corresponding manuscript set constructed by reading the manuscript text set in advance in the corresponding manuscript set storage unit. The corresponding document set represented by a weighted finite state transducer having branches as a network is stored. Then, the speech recognition error correction device calculates, as node data, a score of a state capable of transitioning on the network of the weighted finite state transducer at each time when the input of the word of the recognized word string is received by the node data updating unit. The node data storage means stores the calculated node data at each update time. Then, the speech recognition error correction apparatus performs a predetermined process start condition without waiting for input of recognition results of all recognized word strings for all originals for determining the final best hypothesis by the original searching means. In addition, while tracing back on the network based on the node data stored at that time, a hypothesis partially approximating the final best hypothesis is sequentially determined as an error correction result. Then, the speech recognition error correction apparatus sequentially outputs the corresponding original document determined as the error correction result by the original output unit.

  According to the present invention, it is possible to reduce the occurrence of automatic correction errors caused by block boundaries in the conventional block matching method.

1 is a block diagram schematically showing a system including a speech recognition error correction device according to an embodiment of the present invention. It is a figure which shows typically the construction example of a weighted finite state transducer. It is a block diagram which shows typically the structure of the speech recognition error correction apparatus which concerns on embodiment of this invention. It is FIG. (1) for demonstrating the trace back and original division by the speech recognition error correction apparatus which concerns on embodiment of this invention. FIG. 10 is a diagram (No. 2) for explaining traceback and document division by the speech recognition error correcting device according to the embodiment of the present invention; It is a flowchart which shows the flow of the process by the speech recognition error correction apparatus which concerns on embodiment of this invention. It is a schematic diagram which shows the example of the algorithm which can be applied with a weighted finite state transducer. It is a schematic diagram for demonstrating correction of a recognition error in a prior art. It is a schematic diagram for demonstrating lack of a recognition result in a prior art. It is a schematic diagram for demonstrating the double output of a recognition result in a prior art.

Hereinafter, the speech recognition error correction apparatus of the present invention will be described in detail.
The speech recognition error correction apparatus 100 shown in FIG. 1 accepts as input a word string (recognition word string) of a recognition result output by the speech recognition apparatus 220 that recognizes the uttered speech read out from the document 201 included in the document text set 200. The error contained in the recognized word string is corrected by estimating the word string of the corresponding document stored in advance. Here, the information stored in advance for the estimation process by the speech recognition error correction apparatus 100 is a set of corresponding manuscripts constructed by reading the manuscript text set 200 in advance, and the state transition between nodes representing the state Is a corresponding manuscript set represented by a weighted finite state transducer (hereinafter referred to as WFST) having a branch representing the network as a network. This speech recognition error correction apparatus 100 sequentially checks the best hypothesis on the WFST network and does not wait for the input of all words based on the edit distance between the word string of the corresponding manuscript on the WFST and the recognized word string. The best hypothesis is approximated and the correction results are partially determined sequentially.

  The example shown in FIG. 1 schematically shows the entire system including the speech recognition error correction apparatus 100 to be applied when subtitles are given to a news program originated from a local broadcasting station using speech recognition. The utterance voice of such a program is characterized in that it is generally based on a manuscript text prepared in advance. In addition, large-scale key broadcasting stations have an operator for manually correcting a recognition error included in the speech recognition result. However, it is difficult for local broadcasting stations to arrange the operator at present. An example of this is shown. According to the present embodiment, it is possible to solve the problem of local station operator arrangement.

  A manuscript text set 200 shown in FIG. 1 represents the whole of a text that is a transcript of what a person plans to speak. The manuscript text set 200 is divided into a number of subdivided individual contents, for example, according to word string delimiter units such as sentences, sentences, paragraphs, and content classifications such as themes and topics. Such individual contents are hereinafter simply referred to as a manuscript. Further, description will be made assuming that the word string unit is a sentence as an example.

In the present embodiment, for example, the following conditions (A1) to (A7) are assumed.
(A1) A plurality of manuscript sentences in the manuscript text set 200 are read as voice recognition targets.
(A2) For example, even if the manuscript is related to one news item, several updated versions (versions) are prepared, and it is not known in advance which version of the manuscript will be read in the news program. .
(A3) The order in which a plurality of manuscript sentences are read is not known in advance.
(A4) Some original texts included in the original text set 200 cannot be read.
(A5) Depending on the person reading, there is a case where the wording is intentionally changed or a wording error occurs without reading the document as it is.
(A6) A major premise is to avoid sending a subtitle that has become unclear due to a recognition error of the voice recognition device 220 to mislead or make the viewer uncomfortable. Therefore, it is not sent in the case of an unknown recognition result, and instead, the subtitle is a manuscript (prior manuscript) that is automatically estimated to be the closest to the utterance content that has been proofread by the editor and confirmed in advance. Send it out.
(A7) If there is no original corresponding to the recognition result in an interview part or the like, automatic estimation is impossible, so no subtitles are transmitted for an interview part without an original.

  The manuscript text set 200 is a set of manuscript electronic data submitted by a reporter for, for example, a news program, and is stored in a general storage device such as a hard disk or a storage means on a network. This document text set 200 is also used to construct a WFST of the corresponding document set in advance.

  When raw speech data is input, the speech recognition device 220 recognizes speech data using an acoustic model or language model based on a hidden Markov model (HMM), and generates the recognized result as a recognition word string. To do. In the present embodiment, the voice recognition device 220 is not particularly limited, and a conventionally known device can be adopted.

  As shown in the condition (A2), a plurality of versions of a manuscript have been submitted for each news item, and it is not possible to determine in advance which version will be broadcast in which order. In such a situation, the speech recognition apparatus 220 must perform speech recognition and immediately check whether or not a corresponding document exists for the uttered speech. Therefore, the language model used for speech recognition is adapted using the document text set 200 in order to obtain a high-accuracy correspondence between the speech recognition result and the document, and the recognition accuracy when it is read out as it is according to the document becomes high. It is preferable to do so.

The transducer construction device 240 constructs a WFST as a corresponding document set (corresponding document set) used in the speech recognition error correction apparatus 100.
The transducer construction device 240 constructs in advance a WFST to be used by the speech recognition error correction device 100 from a read-out original to be speech-recognized, that is, from a document sentence included in the document text set 200. WFST is a finite state machine having input symbols, output symbols, and transition weights, and can efficiently handle input / output of different granularities such as words and sentences. The construction of this WFST will be described later.

  Each time a recognition result word is input from the speech recognition device 220, the speech recognition error correction device 100 uses WFST to obtain a transition that can accept the input word, calculates its score, and sets a threshold for the cumulative score. It is assumed that a search (Viterbi search) using a conventionally known Viterbi algorithm is used while performing pruning. Note that the Viterbi algorithm is a method that uses a trellis diagram to effectively calculate a code sequence having the maximum likelihood when estimating a code sequence that is closest to a transmission code with respect to a received sequence, that is, that maximizes the likelihood. It is a method of searching.

  In normal Viterbi search, after all the inputs are observed, the path with the best score is traced back and the best hypothesis is output. Therefore, in the normal search method, it is not possible to output the correction results sequentially from the oldest input before all the inputs are observed. For example, assuming that the subtitles are produced from the result of recognizing the broadcast sound of a TV broadcast program and superimposed on the image of the TV screen in real time, the maximum likelihood sequence by the normal Viterbi search must input words until the end of the program It cannot be confirmed. This will end the program, so normal Viterbi search is not suitable for such operations.

  On the other hand, the speech recognition error correction apparatus 100 successively approximates the maximum likelihood sequence and traces back using the Viterbi search. That is, every time a predetermined processing start condition is satisfied, a path with the best score at that time is traced back and an output transition that can be determined is determined, so that correction results can be output sequentially. The path traced back here is an approximation of the best hypothesis, but whether or not the path is determined based on the reliability of the edit distance between the input word string corresponding to each output transition and the word string of the manuscript. To improve the approximation accuracy. Details will be described later.

[Example of constructed WFST]
FIG. 2 is an example of WFST constructed by the transducer construction device 240. The WFST has a node representing a state and a branch representing a state transition. Note that state transition may be simply referred to as transition. In the present embodiment, a WFST is constructed in which an input symbol is a word and an output symbol is a predetermined word string. A predetermined word string will be described as a sentence.

  In this example, each ellipse node is given a three-digit number for identification. The start point node is node 001, and the end point node is node 008. In this example, nodes 002 to 007 are arranged in a straight line between the start point and the end point. Further, nodes 010 to 015 are arranged in a straight line in parallel between the start point and the end point. Furthermore, nodes 018 to 023 are arranged in a straight line in parallel between the start point and the end point. In this example, a transition (branch) is set between the state (node) and the state (node). Here, when it is between nodes, it also includes between own nodes. Each transition has a word or symbol <S>, <I>, <D>, <EmiX (where X is one of 1 to 3)> and <eps> Either is listed.

First, generalizing and explaining all the transitions in FIG. 2, in this WFST, a parameter (S ⁱ / S ^o : ω) is set for each transition between states. Here, S ⁱ represents a word input accepted by the transition, S ^o represents a predetermined word string (sentence) output by the transition, and ω represents a transition weight (state transition weight). That is, a triple parameter is set for each transition. However, in FIG. 2, the parameters are not described in all transitions for the sake of space, and a total of 18 transitions including words are included in any of the three parameters of S ⁱ and S ^o . Only is described.

  Here, the word described in FIG. 2 is generalized and expressed as a word s. Note that uppercase and lowercase letters are distinguished. In FIG. 2, a word s represents a word included in the word string of the document. Each transition in which the word s is described indicates that transition is possible only when the same word as the word s described in the transition is input. That is, if the word input at the position of the recognized word string corresponding to the position of a certain word s included in the word string of the document is the same as the word s on the document, the state can be changed. In short, each transition in which the word s is described is a transition that proceeds after receiving a speech-recognized word. As described above, the WFST constructed here is a network that can freely connect all document texts.

  In FIG. 2, the parameter for the transition in which the word s is described is represented by (s / ε: 0.0). Here, s represents a word input that the transition can accept, and ε means that there is no output in this transition. Moreover, 0.0 is one of the transition weights, and means that no penalty is imposed when the same word as the word s is input for this transition. For example, the transition in which “last month” is described in FIG. 2 is (last month / ε: 0.0) in terms of a triple parameter.

  In FIG. 2, a transition in which <S> is described is a transition for accepting a replacement word. That is, when the word input at the position of the recognized word string corresponding to the position of a certain word s included in the document word string is replaced with an arbitrary word different from the word s on the document, the replacement is performed. A transition for accepting a word. Hereinafter, an arbitrary word different from the word s at the position of a certain word s included in the word string of the document is referred to as an arbitrary word *. This replacement includes, for example, a case where “restart” is recognized as a transliteration of the homonym “reunion”.

In FIG. 2, the transition in which <S> is described can accept any word *. The parameter for the transition in which <S> is described is represented by (* / ε: ω _s ). Here, * represents an arbitrary word input that the transition can accept, and ε means that there is no output in this transition. Further, ω _s is one of transition weights, and means a penalty imposed on the transition when an arbitrary word * different from the word s is input (hereinafter referred to as a replacement penalty). This replacement penalty ω _s is expressed by a numerical value that lowers the node score, and for example, −1.0 is used. For example, a transition in which <S> is described in FIG. 2 is (* / ε: −1.0) in terms of a triple parameter.

  In FIG. 2, a transition in which <I> is described is a transition for accepting an insertion word. In other words, when there is a repetition of information added to the utterance content or segmentation due to stagnation due to the speaker, the word inserted at the position following the word string recognized as being replaced or replaced is accepted. Transition. In addition, due to the voice recognition device 220, a word inserted in a position following the one word as the manuscript is caused by a recognition error that recognizes a word that should be recognized as the manuscript as a plurality of words if it is as the manuscript. Is a transition to accept

In FIG. 2, the transition with <I> described can accept any word *. The parameter for the transition in which <I> is described is represented by (* / ε: ω _i ). Here, * represents an arbitrary word input that the transition can accept, and ε means that there is no output in this transition. Further, ω _i is one of transition weights, and means a penalty imposed when an arbitrary word * is input for this transition (hereinafter referred to as an insertion penalty). This insertion penalty ω _i is represented by a numerical value that lowers the node score, and for example, −1.0 is used. For example, the transition in which <I> is described in FIG. 2 is (* / ε: −1.0) in terms of a triple parameter.

  In FIG. 2, a transition in which <D> is described is a transition for accepting a dropped word. That is, this is a transition for specifying the position of a word dropped from the original in the recognized word string when a phrase or the like is dropped in a part of the utterance content due to the speaker. In addition, due to the voice recognition device 220, a word that should be recognized as a plurality of words according to the original is caused by a recognition error in which the word is deleted and recognized as one word, and is dropped from the original in the recognized word string. It is a transition for specifying the position of a word.

In FIG. 2, transitions with <D> can be transitioned even if no word is input. The parameter for the transition in which <D> is described is represented by (ε / ε: ω _d ). Here, the first ε means that there is no word input in this transition, and the next ε means that there is no output in this transition. Further, ω _d is one of transition weights, and means a penalty imposed when a word is dropped in this transition (hereinafter referred to as a drop penalty). This drop penalty ω _d is represented by a numerical value that lowers the node score, and for example, −1.0 is used. For example, the transition in which <D> is described in FIG. 2 is (ε / ε: −1.0) when represented by a triple parameter.

  In FIG. 2, a transition in which <EmiX> is described is a transition for outputting a sentence L as a predetermined word string, and a transition for outputting a correction result. The parameter for the transition in which <EmiX> is described is represented by (ε / L: 0.0). Here, ε means that no word is input in this transition. L means a word string (sentence) output in this transition. For example, a transition in which <Emi1> is described in FIG. 2 is (ε / in the Kanto Koshin region in the last month ...: 0.0) in terms of a triple parameter. That is, in this case, L is a word string in which all of the word strings “Kanto Koshin region in last month are ...” arranged in each transition from the start node 001 to the node 007 via the node 002 are connected in order. The parameter 0.0 is one of transition weights, and means that no penalty is imposed when a sentence is output for this transition.

In FIG. 2, the transition in which <eps> is described is a transition that connects the end point node and the start point node, and is called an epsilon transition (ε transition). The transition in which <eps> is described is a transition that gives a constraint that a predetermined word string (sentence) included in the document text set is continuously spoken. A parameter for a transition in which <eps> is described is represented by (ε / ε: ω _u ). The first ε means that there is no word input at this transition, and the next ε means there is no output at this transition. Also, ω _u is one of transition weights, and by giving an appropriate weight (numerical value), the WFST can increase the score of sentences that match longer.

[How to build WFST]
A method for constructing a WFST by the transducer construction device 240 will be described.
A word string unit for arranging output transitions (transitions in which <EmiX> is described) in WFST is determined in advance. This can be set according to the required error correction capability. The positions where the output transitions are arranged can make the unit of the manuscript included in the manuscript text set 200 as one delimiter. As the position where the output transition is arranged, a sentence unit, a phrase unit, a line feed unit arranged for easy reading by the reporter, or the like can be used. Here, when a long unit is set, the correction accuracy increases, but the determination of the subtitle word string to be transmitted is delayed. Conversely, if a short unit is set, the determination of the subtitle word string to be sent out becomes faster, but the correction accuracy decreases. Therefore, what unit should be used may be appropriately designed according to the expected recognition accuracy of voice recognition and the degree of matching between the original and the reading voice.

  In this embodiment, the position where the output transition is arranged in the WFST is determined in units of sentences as an example. From another viewpoint, the WFST in FIG. 2 uses a branch (word s in FIG. 2) that represents input transition of each word constituting a sentence between a start point node 001 and an end point node 008 for each sentence (a word string in a predetermined unit). ) And a branch representing an output transition (transition in which <EmiX> is described in FIG. 2).

  The construction of the WFST starts with the starting point of the WFST, and each time the original text included in the original text set 200 is read word by word, a transition with a weight of 0 for accepting the word and a new node are sequentially created. . Here, the transition of weight 0 means (s / ε: 0.0) when expressed by a triple parameter. And when it becomes the above-mentioned predetermined unit, an output transition is added and it connects with the end point node of WFST. If the original still remains, starting from the starting point again, each time the original text is read word by word, a transition with a weight of 0 and a new node for accepting the word are sequentially created. And when it becomes the above-mentioned predetermined unit, an output transition is added and it connects with the end point node of WFST. Thereafter, the same is repeated.

When all the original texts have been read from the original text set 200, finally, the end point node and the start point node are connected by the ε transition. Here, the ε transition is (ε / ε: ω _u ) in terms of a triple parameter. Here, an appropriate weight is given to the transition weight ω _u . As a result, the WFST can increase the score of a sentence that matches longer, and can operate properly even when a sentence that matches the prefix of another sentence exists in the document. Finally, transitions that accept substitutions, omissions, and insertions are added to the transitions of each word.

[Configuration example of transducer construction device]
In the example shown in FIG. 1, the transducer construction device 240 includes a word network registration unit 241 and an editing network registration unit 242.
The word network registration unit 241 performs the following series of processes for each predetermined unit (for example, sentence unit) in the document text included in the document text set 200. That is, as a series of processes, the word network registration unit 241 receives input words from the start node of the WFST network each time a word in a word string included in the original text included in the original text set 200 is read. Branches and new nodes are sequentially created until the read word string reaches a predetermined unit (for example, sentence unit). Then, in the WFST network, an output transition branch of the read word string is added and connected to the end node.

  The editing network registration unit 242 corresponds to a branch representing a state transition that accepts an arbitrary word corresponding to a word replacement and a word insertion between nodes of the WFST network created by the word network registration unit 241. Thus, a branch representing a state transition that accepts an arbitrary word and a branch representing a state transition that transitions to the output side even if there is no input in response to the deletion of the word are added.

[Configuration example of speech recognition error correction device]
In the example shown in FIG. 1, the transducer construction device 240 is provided separately from the speech recognition error correction device 100. However, as shown in FIG. 3, for example, the speech recognition error correction device 100 may include the transducer construction device 240. Good. As shown in FIG. 3, the speech recognition error correction apparatus 100 includes a WFST storage unit (corresponding document set storage unit) 110, a node data update unit 120, a node data storage unit 130, a document search unit 140, and a document. Output means 150.

  The WFST storage unit (corresponding document set storage unit) 110 stores a WFST (corresponding document set) constructed in advance using the document text set 200. This WFST (corresponding document set) is constructed by the transducer construction device 240. Therefore, since WFST is the same as that described with reference to FIG. 2, description thereof is omitted to avoid duplication.

  The node data updating unit 120 calculates and updates a score of a state that can be transited on the WFST network as node data every time when an input of a word of a recognized word string output from the speech recognition device 220 is received. For example, each time a recognition word is input, the node data updating unit 120 refers to the WFST stored in the WFST storage unit 110 and sequentially performs a Viterbi search to update the node data.

The node data updating unit 120 adds “0” to the score when the word input as the recognition word string is the same word as the corresponding document, and penalizes the score when the input word is a word different from the corresponding document. Add "-1".
For example, in the example shown in FIG. 2, when the word string input as the recognition word string is exactly the same word string as the corresponding document, the word “Last Month” is received from the start node 001 and corresponds to the word of the corresponding document. Since the process proceeds to the node 002 through the transition, the node data update unit 120 adds “0” to the score. Thereafter, for example, when “NO” is received and the process proceeds to node 003, “0” is added to the score. Similarly, when “Kanto Koshin”,... Is received, “0” is added to the score.

  On the other hand, for example, in the example shown in FIG. 2, when the word string input as the recognition word string is a word string different from the corresponding manuscript, if the word “Last Week” is received from the start point node 001, ”Has been replaced, the process proceeds to node 002 through a transition corresponding to the replacement. In this case, the node data updating unit 120 adds a penalty “−1” to the score. Also, when passing through a transition corresponding to an insertion error or dropout error, the node data update unit 120 similarly adds a penalty “−1” to the score.

  Thus, when the input recognition word is the same as the word s in WFST, the score of the path is the best. On the other hand, if there is a replacement, insertion, or deletion edit, the score deteriorates. For example, a transition in which <D> is described can transition even if there is no input. However, in the case of a path that passes only a transition in which <D> is described, the score becomes lower as it approaches the output transition. The WFST is created as a network in which if the recognition word string includes an error or paraphrase, the score is deteriorated accordingly.

  The node data storage unit 130 stores the node data calculated by the node data update unit 120 at each update time, and is a general storage unit such as a memory or a hard disk.

  The document searching means 140 is stored at that time every time a predetermined processing start condition is satisfied without waiting for input of recognition results of all recognized word strings for all documents for determining the final best hypothesis. A hypothesis partially approximating the final best hypothesis is sequentially determined as an error correction result while tracing back on the WFST network based on the node data.

  The document search unit 140 approximates the final best hypothesis based on the edit distance between the word string of the corresponding document included in the WFST (corresponding document set) and the input recognition word string. The document search means 140 determines that the path section is reliable if the edit distance in the path section from the head to the end is small to some extent between the paths divided in predetermined ranges on the WFST network. And output. Here, the short edit distance means that the path through which the recognized word string and the original word string are almost matched has been passed. On the other hand, since the reliability of a path section with a long editing distance is low, it is not determined at that time and is used for the next traceback. It is presumed that a path section having a low reliability forever is a section talking about a difference that is not originally described in the manuscript. Therefore, a path section with low reliability is not output. In the following, a path section in a predetermined range on the WFST network will be described as an example of a path section sandwiched between two output transitions on the WFST network.

  The processing start condition is satisfied when, for example, a silent period without speech is reached a predetermined period, or the number of input words as a recognized word string output by the speech recognition device 220 is a predetermined number of words. It means the case of reaching. The predetermined period is not particularly limited, but an example is 3 seconds. Further, the predetermined number of words is not particularly limited, but 20 words can be given as an example. The activation signal at this time may be automatically output to the voice recognition device 220, for example, or manually when the operator recognizes that it is a pose or when a predetermined number of words has been reached. You may make it input. According to this, processing load can be reduced compared with the case where a search process is started for every input of a recognition word. Further, for example, if the silent period is only a predetermined period, since the sequential reception of the recognition results is stopped during that period, the node scores at that time can be easily compared.

  In order to realize the above function, in the present embodiment, the document search unit 140 includes a maximum score node detection unit 141, a traceback unit 142, a document division unit 143, an output candidate, as shown in FIG. The storage unit 144, the edit distance calculation unit 145, the edit distance determination unit 146, the fixed output storage unit 147, and the fixed time storage unit 148 are provided.

  The maximum score node detection unit 141 detects a node having the maximum score in the node data stored at that time when a predetermined process start condition is satisfied. For example, when the silent period (pause) in which there is no uttered voice reaches a predetermined period, or whenever the number of input words as a recognition result reaches a predetermined number of words, the activation signal indicating that is the maximum score Input to the node detection means 141.

  The traceback means 142 follows the WFST network from the downstream to the upstream for the path that has reached the node from the node detected by the maximum score node detection means 141, and is confirmed by the previous traceback, and the output word Trace back to the time corresponding to the last input word of the sequence.

  FIG. 4 is a schematic diagram in which a path P1 is added to the WFST shown in FIG. In FIG. 4, it is assumed that the node with the highest score is the node 020. Further, it is assumed that the node corresponding to the last input word determined in the previous traceback is the node 007. In this case, the traceback unit 142 traces the path P1 from the position indicated by the star in the reverse direction in the order of the node 020, the node 019, and the node 018, and when it reaches the start point node 001, returns to the end point node 008. Next, the node 015 is reached through the output transition <Emi2> of the second tree. Subsequently, the traceback unit 142 traces backward in the order of the nodes 015, 014,... In FIG. 5, and when reaching the start point node 001, returns to the end point node 008 as shown by the path P2. Next, the node 007 is reached through the output transition <Emi1> of the first tree.

Returning to FIG. 3, the description of the document search means 140 will be continued.
The document dividing means 143 cuts out a word string of the corresponding document included in the WFST (corresponding document set) for each path section sandwiched between two output transitions in the path traced back this time. In the case of the example described with reference to FIGS. 4 and 5, the path section sandwiched between the output transition <Emi1> and the output transition <Emi2> is divided by the document dividing unit 143.

  The output candidate storage unit 144 stores an output transition output symbol (cut out document) corresponding to the path section divided by the document dividing unit 143 as an output candidate. Storage means. In the case of the example described with reference to FIGS. 4 and 5, “This week's rain is ...” is stored as an output candidate.

The edit distance calculating unit 145 calculates an edit distance from the input recognition word string for each corresponding document cut out by the document dividing unit 143. In the present embodiment, the editing distance is defined by a value obtained by dividing the number of editing operations related to insertion, replacement, and deletion for the path section by the number of words in the path section. Here, assuming that the number of word replacement, insertion, and deletion editing operations in the recognized word string is e, and the number of original words corresponding to the output transition is N _r , the editing distance is recognized for the original word number N _r . It is represented by the ratio (e / N _r ) of the number of word editing operations e in the word string.

  Specifically, in the example shown in FIG. 2, the case where the path section on WFST is “node 007 → node 008 → node 001 → node 0010 → node 0011 → node 0012 → node 0013 → node 0014 → node 0015”. Suppose. This path section is composed of 6 words, and when the word “this week” is replaced with “this month” and recognized, the edit distance is 1/6.

The edit distance determining means 146 sequentially determines whether or not the calculated edit distance is equal to or less than a predetermined threshold while selecting a path section from the downstream to the upstream in the WFST network. The output transition of the path section on the WFST network is determined, and the output symbol is determined as an error correction result. Here, when the edit distance (e / N _r ) is equal to or smaller than the threshold T, that is, when e / N _r ≦ T is satisfied, the output symbol of the output transition is determined as the correction result. Further, when the edit distance (e / N _r ) is larger than the threshold value T, the edit distance determination unit 146 does not adopt the output symbol. That is, the output of the output transition of the path section having an edit distance larger than the threshold is temporarily suspended, and is rejected when there is an output transition determined after this path section. Since the edit distance (e / N _r ) is a value in the range of 0 to 1 from the definition, the threshold satisfies the relationship 0 <T <1.

  The definite output storage unit 147 stores the output symbol of the output transition in the predetermined path section as an error correction result when the editing distance determination unit 146 determines that the editing distance is equal to or less than the predetermined threshold. For example, it is a general storage means such as a memory or a hard disk. The storage structure of the definite output storage unit 147 is a stack and holds data in a last-in first-out structure.

  The fixed time storage unit 148 stores the fixed time determined by the current traceback process, and is a general storage unit such as a memory or a hard disk. The confirmation time storage means 148 is the latest confirmation corresponding to the output symbol stacked on the stack at the time when the edit distance determination means 146 has finished the determination for all the path sections (all cut out documents) to be traced back this time. The time of the word is stored as the confirmed time.

  The manuscript output unit 150 sequentially outputs the corresponding manuscript determined as an error correction result by the manuscript search unit 140. The manuscript output means 150 is determined by the end of the determination process for the edit distance calculated for each path section of all the cut out corresponding manuscripts in the path traced back through the WFST network this time. The data of the output symbols loaded on the stack is output until the stack becomes empty.

  The correction output by the speech recognition error correction apparatus 100 includes both the meaning of correcting an error and not outputting the error. In other words, if the correction result by the speech recognition error correction device 100 can be viewed in advance by a person, an error is enough to feel that “this is not a sentence” or “the meaning is different”. The operation in which the speech recognition error correction apparatus 100 detects the part in the process and does not output the detected part is included as error correction in a broad sense.

[Operation of voice recognition error correction device]
The flow of processing by the speech recognition error correction apparatus 100 according to the embodiment of the present invention will be described with reference to FIG. 6 (refer to FIG. 3 as appropriate).
(Assuming 1) the recognition result of a word input _{{ω 0, ω 1, ...} , ω k, ..., ω j, ...} and.
(Assumption 2) _Let ω _k be the last input word of the part determined by the previous traceback, and let the output transition at that time be a _p (Pth output transition along the time axis).
(Premise 3) Consider a case in which, after a word ω _{j as a} recognition result is input, successive determination is performed using a predetermined silence as a trigger.
(Premise 4) The node data updating unit 120 accepts the last input word ω _j before silence and calculates all nodes that can make a transition.

The maximum score node detecting means 141 detects the node having the highest score in the node data stored at the present time, triggered by the predetermined silence being continued (step S1). The state represented by this detection node is the maximum likelihood state at the start of traceback. Then, the traceback unit 142 traces the word history on the WFST in the reverse direction for the path reaching the node from the detected node, confirms it in the previous traceback, and outputs the last input word of the output word sequence. Trace back to a fixed time corresponding to ω _k (transition in which the word accepted by the WFST transition is ω _k ) (step S2). Here, the confirmed time stored in the confirmed time storage means 148 is used as the confirmed time corresponding to the last input word ω _k of the word sequence that is confirmed and output in the previous traceback. Instead of the transition whose word is ω _k , traceback may be performed until the output transition a _P is reached.

Then, the document dividing unit 143 divides the document for each path section sandwiched between two output transitions in the path traced back this time, and stores the document in the output candidate storage unit 144 as an output candidate (step S3). . Here, the document is divided every time it passes through the output transition a _L (L-th (but L> P) output transition along the time axis) that can be output while proceeding in the reverse direction until reaching the output transition a _P. Alternatively, the document may be divided every time it passes through the output transition a _L that can be output from the output transition a _P side. The output transition a _L that can be output includes output transition symbols that are candidates for output, but also include output transitions that are later rejected by the edit distance determination unit 146 and are not output. The edit distance of such an output candidate is denoted as D.

Then, the edit distance calculation unit 145 calculates the edit distance D of the output candidate (step S4). Specifically, the section that corresponds to the output symbol of the output transition a _L, i.e., immediately before the output transition a _L-1 and with the output transition a _L of when it proceeds in the opposite direction on WFST from the output transition a _L A value obtained by dividing the number of editing operations (that is, the number of passes through <S>, <D>, and <I>) for the path section sandwiched between by the number of words in the section is the output transition a _L Is calculated as the edit distance D _L at. That is, assuming that the number of editing operations in the same section is e _L and the number of words in the same section is N _L ^r , the editing distance D _L in the output transition a _L is expressed by e _L / N _L ^r .

  Then, the edit distance determination unit 146 selects a predetermined output candidate, and determines whether or not the calculated edit distance D is equal to or less than the threshold value T (step S5). When the edit distance D is less than or equal to the threshold T (step S5: Yes), the edit distance determination unit 146 determines the output transition of the path section on the WFST and determines the output symbol as an error correction result (step S6). ). Further, the editing distance determination unit 146 accumulates the data of the output symbol determined this time on the stack stored in the finalized output storage unit 147 (step S7), and proceeds to step S8.

  Then, when there is an output candidate that should still be selected ahead (Step S8: No), the edit distance determination unit 146 returns to Step S5. On the other hand, when all the output candidate selections are selected (step S8: Yes), that is, when the determination process for the edit distance calculated for each path section corresponding to all the cut out documents is completed, The document output means 150 sequentially outputs the output symbol data stacked on the stack at that time until the stack becomes empty (step S9). As a result, the documents are output sequentially from the document placed on the front side.

  Here, when all the output candidate selections are selected (step S8: Yes), the edit distance determination unit 146 selects the latest confirmed word corresponding to the output symbol stacked on the stack for the current traceback process. Is stored in the fixed time storage means 148 as the fixed time determined in step.

  If the edit distance D is greater than the threshold T in step S5 (step S5: No), the process proceeds to step S8 without loading data on the stack.

  That is, the document output unit 150 sequentially outputs the data accumulated on the stack in each traceback process as a confirmed document. At this time, if the edit distance D of the predetermined path section is larger than the threshold T in the speech recognition result, the output symbol of the output transition of the path section is adopted as the error correction result because the path has low reliability. And no output.

[How to determine the edit distance threshold T]
If the recognition accuracy of voice recognition is about 90%, the edit distance value may be about 90%. The threshold value T of the edit distance used for the determination is preferably set at a position sufficiently lower than the recognition accuracy of voice recognition, for example, with a margin below the reliability of the word matching rate. Here, the reliability of the word match rate depends on the number of words between two output transitions of the WFST network.

As another factor, it is preferable to determine the threshold value T in consideration of the ratio of the overlapping of the original document candidates included in the original text set 200 as a ratio. For example, in the case of sentences shown in (E1) to (E3) below, overlapping as sentences is included at a rate of about 80%.
(E1) Today's weather is sunny (E2) Today's weather is rainy (E3) Today's weather is cloudy In this case, if the edit distance threshold is set to about 80%, the desired movement Cannot be realized. It has been confirmed that when an experiment is performed under the condition that an output transition is arranged for each sentence of the news manuscript and the threshold T is 50%, the operation is performed without any problem.

[WFST options]
<Option 1: Construction of WFST that accepts paraphrasing>
A manuscript serving as a WFST information source may include a phrase that is skipped when it is read, a phrase that can be rephrased, and a phrase that is supplemented. Some of these are typical and occur frequently. For example, in a news program manuscript, phrases such as “according to the Metropolitan Police Department” representing the source of the interview are fixed phrases that are easily skipped. However, even if this is skipped, there is no change in the meaning of the news main sentence (5W1H), and there is no practical problem.

  In option 1, such a standard wording is added to the WFST so that the correction result can be output with high accuracy. As conventionally known, WFST is used for speech recognition decoders, machine translation, and the like, and various calculation algorithms are known. For example, an algorithm that performs synthesis (see FIG. 7A), minimization (see FIG. 7B), and determinization (see FIG. 7C) can be applied, and an efficient state transition machine can be configured. There is a feature that you can. The addition of the wording can be efficiently realized by separately constructing a WFST for adding wording separately from the WFST constructed from the manuscript and combining it with the WFST constructed from the manuscript.

  For example, as for paraphrasing examples, prepare the ones that have a high frequency and that have no change in meaning by paraphrasing, based on the difference between the original of the same kind of program in the past and the word string actually read out. deep. For each selected paraphrase example, a WFST for synthesizing paraphrases is constructed, and a WFST that is compatible with paraphrasing can be constructed by performing a synthesis operation with the WFST constructed from the original. Here, the synthesis of WFST will be described with reference to FIG.

  In FIG. 7A, the nodes are shown as circles. The upper diagram on the left side of FIG. 7A is a schematic diagram of an example of a WFST constructed from a document, and the lower diagram on the left side of FIG. 7A is a schematic diagram of an example of an added WFST. . The diagram on the right side of FIG. 7A is a schematic diagram of the WFST after the WFST constructed from the original and the added WFST are combined.

<Option 2: Option A when creating WFST>
When creating the WFST, the WFST may be minimized if necessary. Here, the minimization of WFST will be described with reference to FIG. The diagram on the left side of FIG. 7B is a schematic diagram showing an example of a WFST constructed from a document by a normal method. Here, a1 to a6 indicate different words.

The diagram on the right side of FIG. 7B is a schematic diagram of the WFST after minimizing the WFST constructed from the original by a normal method. In the WFST after minimization, nodes (states) are aggregated in consideration of the arrangement order (word position) for the prefixes (words a1, a2) common to the three word strings of the original WFST, and the minimum number Branches (transitions) are arranged.
According to WFST minimization, word strings (sentences) having the same prefix can be shared by the same transition, so that the amount of calculation can be reduced.

<Option 3: Option B when creating a WFST>
Further, when creating a WFST, if necessary, the WFST may be determinized. Here, determinization of WFST will be described with reference to FIG. The diagram on the left side of FIG. 7C is a schematic diagram of a WFST having the same shape as the WFST shown on the right side of FIG. However, in the transition in which the word a4 is described in FIG. 7B, the output sentence o1 is described instead. Similarly, in the transition in which the word a5 is described, the output sentence o2 is described instead, and the output sentence o3 is described instead of the word a6.

The diagram on the right side of FIG. 7C is a schematic diagram of the WFST after determinating the original WFST. The WFST after determinization is different from the original WFST in that the output sentence o3 is described in the previous transition (one left side).
The original WFST has a branch at the time of state transition from the second node from the left to the next node, and when the transition from the second node to the lower node in the figure, the output statement is o1 or o2. It turns out that it becomes decisive to become o3. Therefore, in order to output the estimation result as soon as possible, the position of the output sentence is changed in the WFST after determinization.

  According to WFST determinization, there is an advantage that the output sentence is moved to a transition with a unique prefix, so that the output sentence can be finalized. However, if the WFST is determinized when creating the WFST, it is necessary to change the setting so that the search process of the maximum likelihood hypothesis by the manuscript search means 140 can also cope. That is, it is necessary to shift the path section for calculating the edit distance before and after the output transition as compared with the case where WFST is not determinized. In addition, the threshold value T needs to be set to a stricter value (small value) so that the expansion and contraction of the preceding and following path sections can be absorbed.

[Other options]
The present invention can also be applied to the generation of multilingual subtitles. For example, in the output transition in which <Emi2> next to node 015 of WFST shown in FIG. 2 is described, an English sentence corresponding to a Japanese sentence from nodes 010 to 015 is used as an output symbol, thereby supporting Japanese voice input. English subtitles can be generated. If you need to generate Japanese and English subtitles at the same time, you can use the English translation along with “This rain is all together this week”.

  As described above, the speech recognition error correction apparatus 100 according to the present embodiment recognizes a sentence without fixing a sentence boundary under the constraint that sentences in a document are continuously uttered in an arbitrary order. By taking the correspondence between the result and the manuscript in word units, the automatic correction error of the conventional block matching method is eliminated. On the other hand, in order to obtain a corrected output with higher accuracy, it is desirable that the output has a sentence or a unit equivalent thereto. In order to eliminate this contradiction and achieve compatibility, the speech recognition error correction apparatus 100 uses a weighted finite state transducer (WFST) to determine the correspondence between the recognition result and the document.

  Then, the speech recognition error correction apparatus 100 compares the word string of the recognized word with the word string of the original document to determine where the word string matches the longest of the conventional block matching method (the technique of Patent Document 1). Collating in a range longer than a word chain block of length N (number of words N). Compared to the conventional block matching method, the section for matching the word string of the recognized word and the word string of the manuscript is not only the section corresponding to the word chain block, but the manuscript sentence, which is longer Match the entire sentence. Therefore, it is clearly known where to make the match, and automatic correction errors can be reduced more than before.

  Although the speech recognition error correction apparatus according to the present invention has been described based on the embodiments, the present invention is not limited to these. For example, it was decided by using the edit distance whether or not the reliability of the estimated correspondence manuscript with respect to the recognition word of the utterance speech is high. The insertion rate may be used or they may be used in combination.

  For example, in the output transition in which <EmiX> of WFST shown in FIG. 2 is described, “,”, “.”, Symbols, and the like that cannot be obtained as a result of speech recognition can be embedded in the output symbols according to the notation of the manuscript. . In this case, subtitles that are easier to read can be generated.

  In the present invention, it is not essential to add subtitles. Also, if the content of the speech to be spoken is determined to some extent and the content can be obtained in advance, it is not necessarily premised on the sound of the broadcast program.

In order to confirm the performance of the speech recognition error correction apparatus 100 of the present invention, the following experiment was performed.
The speech recognition target was 420 sentences excluding the weather corner acquired from nine programs of the news program (Metropolitan area news 845). Using these voice recognition results as input, the 420 sentence correction results are visually confirmed and analyzed, and the analysis results are “appropriate / consent”, “unsuitable”, “no output”, “difficult to judge”. It was classified into four types. Here, “inappropriate” represents something that is not established as a sentence or has a different meaning. Further, “difficult to determine” represents paraphrasing over a plurality of sentences, or something that cannot be determined as “agreement”. The analysis results are shown in Table 1.

  As shown in the analysis results in Table 1, the text classified as appropriate / consent was more in the examples. In the comparative example, 18% of the output was inappropriate, whereas in the example, there was no inappropriate output. The reason why there is more “no output” in the embodiment is that, in the embodiment, when the edit distance D of a predetermined path section is larger than the threshold T in the speech recognition result, the path is regarded as a path with low reliability. This is because the output symbol of the output transition of the section is not output. In fact, all the 43 sentences that were not output were those for which no corresponding manuscript originally existed, such as an interview VTR voice.

  Therefore, at least the condition (A6) that is assumed when subtitles are given to a news program originated from a local broadcasting station described in the above-described embodiment by using speech recognition, that is, because of a recognition error of the speech recognition device 220 is significant. It was confirmed that the embodiment was suitable when it was premised that the subtitles that became unknown were sent to avoid misleading or making the viewer uncomfortable.

100 Voice recognition error correction device 110 WFST storage means (corresponding document set storage means)
120 Node data update means 130 Node data storage means 140 Document search means 141 Maximum score node detection means 142 Traceback means 143 Document division means 144 Output candidate storage means 145 Edit distance calculation means 146 Edit distance determination means 147 Final output storage means 148 Confirm Time storage means 150 Document output means 200 Document text set 220 Speech recognition device 240 Transducer construction device 241 Word network registration means 242 Editing network registration means

Claims (6)

Receiving a recognition word string output by a speech recognition device that recognizes a utterance voice read out from a manuscript included in the manuscript text set as an input, and estimating a word string of the corresponding manuscript from a prestored corresponding manuscript set, A speech recognition error correction device for correcting an error included in a recognition word string,
The corresponding manuscript set constructed by reading the manuscript text set in advance and represented by a weighted finite state transducer having a node representing a state and a branch representing a state transition between the nodes as a network Corresponding document set storage means for storing
Node data updating means for calculating and updating a score of a state capable of transitioning on the network of the weighted finite state transducer as node data for each time of receiving an input of a word of the recognition word string;
Node data storage means for storing the calculated node data for each update time;
Each time a predetermined process start condition is satisfied without waiting for input of recognition results of all recognition word strings for all manuscripts for determining the final best hypothesis, based on the node data stored at that time A document search means for sequentially confirming a hypothesis partially approximating the final best hypothesis as an error correction result while tracing back on the network.
Document output means for sequentially outputting the corresponding document determined as the error correction result;
A speech recognition error correction apparatus comprising: The weighted finite state transducer pre-constructed as the corresponding document set stored in the corresponding document set storage means is the network,
For each corresponding manuscript included in the corresponding manuscript set, a branch representing an input transition of each word constituting a word string of the corresponding manuscript and a branch representing an output transition of the word string between a start node and an end node for each corresponding manuscript. Including
A branch representing a state transition from the end node to the start node, and
A branch representing a state transition that accepts an arbitrary word in response to a word substitution, a branch representing a state transition that accepts an arbitrary word in response to an insertion of a word, and no input corresponding to a deletion of a word The speech recognition error correction apparatus according to claim 1, further comprising at least one of a branch representing a state transition that makes a transition to the output side. The document search means includes:
Corresponding manuscript words in a path section in a predetermined range on the weighted finite state transducer network as an edit distance between the word sequence of the corresponding manuscript included in the corresponding manuscript set and the input recognition word string Calculate the value obtained by dividing the number of editing operations related to insertion, replacement, and deletion for a column by the number of words in the path section,
The speech recognition error correction apparatus according to claim 2, wherein the final best hypothesis is approximated by comparing the edit distance calculated for each path section with a predetermined threshold. The document search means includes:
When a word string of a corresponding document whose edit distance is less than or equal to the threshold is determined, the corresponding document whose output is sequentially determined before the corresponding document path section in the weighted finite state transducer network is traced back. Documents corresponding to all path sections whose edit distance is less than or equal to the threshold are sequentially output from the upstream of the network by the document output means, and documents corresponding to all path sections whose edit distance is greater than the threshold are not output. The speech recognition error correction apparatus according to claim 3, wherein The weighted finite state transducer pre-constructed as the corresponding document set stored in the corresponding document set storage means is the network,
A branch that accepts a word string of a predetermined paraphrase candidate having the same meaning as a word string included in the manuscript text set, and / or a word string included in the manuscript text set, and the speech recognition apparatus 5. The apparatus according to claim 1, further comprising a branch that accepts a word string that is predetermined as a possibility of dropping in the recognition word string to be output. Voice recognition error correction device.   The document search means, when the silence period without the utterance voice reaches a predetermined period, or when the number of input words as a recognition word string output by the voice recognition device reaches a predetermined number of words, 6. The speech recognition error correction apparatus according to claim 1, wherein trace back is performed on a network of the weighted finite state transducers assuming that the processing start condition is satisfied.

Images