JP4513165B2 - Program recording method, program recording apparatus, program recording / reproducing apparatus, and program recording / reproducing method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a program recording method, a program recording apparatus, a program recording / reproducing apparatus, and a program recording / reproducing method . More specifically, the present invention relates to a program recording method, a program recording apparatus, a program recording / reproducing apparatus, and a program recording / reproducing method for converting an audio signal recognized when recording a program while recognizing an audio signal to record and store the converted text data.
[0002]
[Prior art]
Conventionally, when a broadcast program is recorded, an EPG (electronic program guide) or the like is searched to confirm the program contents, and a desired broadcast program is recorded. When searching for a broadcast program recorded on a recording medium, a list of recorded broadcast program names is displayed, and a method for selecting a desired program from the list, or the beginning of the recorded program is detected and reproduced. There are methods using the cue function.
[0003]
[Problems to be solved by the invention]
However, EPG information has a standard classification method, and when searching, there is a broadcast program that cannot be searched with a desired keyword, and that does not have an EPG added thereto. On the other hand, when many programs are recorded on the same recording medium, there is a problem that it is impossible to search for a desired broadcast program or a desired scene of the broadcast program, or to perform these searches using arbitrary keywords. .
[0004]
Therefore, there is a problem to be solved in that the recorded broadcast program and a desired scene in the broadcast program can be searched with a keyword input by the user.
[0005]
[Means for Solving the Problems]
In order to solve the above problems, a program recording method, a program recording apparatus, a program recording / reproducing apparatus, and a program recording / reproducing method according to the present invention are configured as follows.
[0006]
(1) Record the program while performing voice recognition,
The voice recognition result is converted into text data, recorded and stored,
Analyzing the content of the program program from the text data, recording and storing the content of the analyzed program program,
A program recording method for searching stored text data using a specific audio signal as a key ,
A program recording method, wherein the appearance frequency of each voice recognition unit is recorded together with the result of the voice recognition.
[0007]
(2) means for recording a program while performing voice recognition;
Means for converting the voice recognition result into text data and recording and storing the result;
Means for analyzing the content of the program program from the text data, and recording and storing the content of the analyzed program program;
Means for searching stored text data using a specific audio signal as a key, and a program recording apparatus comprising:
A program recording apparatus comprising means for recording the appearance frequency of each voice recognition unit together with the result of the voice recognition.
[0008]
(3) means for recording a program while performing voice recognition;
Means for converting the voice recognition result into text data and recording and storing the result;
A program recording / playback apparatus comprising: a playback unit that compares an input keyword with the stored text data and plays back a program related to the matched text data,
A program recording / reproducing apparatus comprising means for recording the appearance frequency of each voice recognition unit together with the result of the voice recognition.
(4) In the program recording / playback apparatus according to (3) above,
A program recording / reproducing apparatus, wherein the voice recognition units are displayed in the descending order of appearance frequency, are appropriately selected, and a program related to the text data of the selected voice recognition units is reproduced.
(5) In the program recording / playback apparatus according to (3) above,
A program recording / reproducing apparatus comprising means for analyzing the contents of a program program from the text data, and recording and storing the contents of the analyzed program.
(6) In the program recording / playback apparatus according to (5) above,
A program recording / reproducing apparatus characterized in that the contents of the stored program program are displayed and selected as appropriate, and a program related to the contents of the selected program program is reproduced.
(7) Record the program while performing voice recognition,
The voice recognition result is converted into text data, recorded and stored,
A program recording / playback method for playing back a program related to matched text data by comparing an input keyword and the stored text data,
A program recording / reproducing method, wherein the appearance frequency of each voice recognition unit is recorded together with the result of the voice recognition.
[0009]
Thus, when recording a program , the audio signal is recognized and recorded and stored as text data, so that the contents of the program can be searched later based on this text data, or the text data can be processed and processed by the user. Secondary use that is presented in the above becomes easy.
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Next, embodiments of a program recording method, program recording apparatus, program recording / reproducing apparatus, and program recording / reproducing method according to the present invention will be described with reference to the drawings.
[0011]
As shown in FIG. 1 , a program recording apparatus, a program recording / reproducing apparatus, and a program recording / reproducing apparatus embodying the program recording / reproducing method according to the present invention are configured to divide a digital broadcast program TS (transport stream) into predetermined packets. Unit 10, a control unit 20 for controlling data to be stored, an audio signal processing unit 30 for processing the result of speech recognition, an information storage unit 40 for storing data relating to a broadcast program including text data, and an information storage unit The MUX unit 50 is a general configuration including a predetermined packet of data stored in 40.
[0012]
The DMUX unit 10 is a demultiplexer, and outputs the digital broadcast program TS (transport stream) separately into image packets, audio packets, time information as a counter value, and program information that is information about other streams.
[0013]
The control unit 20 instructs the information storage unit 40 to record, erase, and output packets according to the program information and externally input information.
[0014]
The voice signal processing unit 30 synchronizes the voice packet and the character data, a DEC unit 31 that restores the voice signal from the voice packet according to the time information, a voice recognition unit 32 that recognizes the voice signal and converts it into character data, and Are composed of a shift unit 33 for generating the time information and a synthesis unit 34 for synthesizing and outputting the character data and the time information.
[0015]
The information storage unit 40 records and stores voice packets and image packets, character data (text data) processed by the voice processing unit 30, program information, and the like in the recording medium 41 under the control of the control unit 20. Further, the storage information designated by the control unit 20 is extracted from the recording medium 41 and output to the MUX unit 50.
[0016]
The MUX unit 50 is a multiplexer, and synthesizes audio packets, image packets, character data, and program information extracted from the information storage unit 40 into a broadcast program stream (TS; transport stream) according to an instruction from the control unit 20. Output.
[0017]
A description will be given of the operation when the result of voice recognition is recorded as text data when a digital broadcast program is recorded by the program recording apparatus having such a configuration, and the recorded data is retrieved and reproduced.
[0018]
First, a digital broadcast signal (MPEG2-TS) received from an external tuner or the like is sent to the DMUX unit 10. The DMUX unit 10 analyzes the digital broadcast signal, and outputs the image packet, the audio packet, time information (counter value) synchronized with the system clock, and program information.
[0019]
The output voice packet and image packet are sent to the information storage unit 40, and the program information is sent to the control unit 20. At the same time, the voice packet and time information are also sent to the voice processing unit 30.
[0020]
The control unit 20 instructs the recording medium 41 to record voice packets and image packets sent to the information storage unit 40 based on the program information output from the DMUX unit 10.
[0021]
On the other hand, the voice processing unit 30 performs voice recognition based on the voice packet and time information from the DMUX unit 10 and converts the result into character data (text data). This process will be described below with reference to FIG. 2 together with FIG.
[0022]
First, the voice packets output from the DMUX unit 10 are voice signal packets A to E having time information T1 to T5, respectively, as shown in FIG.
[0023]
The DEC unit 31 restores the audio signal packets A to E to the baseband audio signals A ′ to E ′ in the order of the time information T1 to T5 and sends them to the audio recognition unit 32 (see FIG. 2B). ). The speech recognition unit 32 recognizes speech signals A ′ to E ′, converts them into character data (text data) a to e, and sends them to the synthesis unit 34. Since it takes a conversion time α to convert the voice signal into character data, the time information of the recognized character data a to e is delayed by the conversion time α with respect to the times T1 to T5 that had before the character conversion. It will be. Since the character data a to e are a collection of words w1 to wn, each of the words w1 to wn has time information of times t1 to tn including a delay of the conversion times α1 to αn (see FIG. 2 (C)) .
[0024]
On the other hand, in the shift unit 33, the speech recognition unit 32 manages the delay of the speech recognition and conversion time α to character data and sends it to the synthesis unit 34 as a counter value. The synthesizing unit 34 synchronizes the voice packets recorded on the recording medium 41 with the character data a to e, so that the time information t1 held by the words w1 to wn of the character data a to e sent from the voice recognition unit 32. New time information of (t1−α1) to (tn−αn) is generated by subtracting the word conversion times α1 to αn (delay) from .about.tn. Then, the time information (t1-α1) to (tn-αn) is added as new time information to the corresponding words w1 to wn (see FIG. 2D).
[0025]
By performing such processing, character data a ′ to e ′ having time information t1 ′ to tn ′ synchronized with the voice packet as shown in FIG. 3 is generated, and the character data a ′ to e. 'Is sent to the information storage unit 40. According to the control unit 20, the information storage unit 40 classifies the image packet, the voice packet, and the program information sent from the DMUX unit 10 and the character data converted by voice recognition into each attribute and records them on the recording medium 41. (See FIG. 4).
[0026]
When a desired keyword is input when searching for the broadcast program stored in this manner, the input keyword is processed by the control unit 20 and sent to the information storage unit 40. The information storage unit 40 compares and refers to the input keyword and the character data stored in the recording medium 41, and searches for matching character data. If there is matching character data, the image packet, sound packet, and program information associated with the time information of the character data are extracted and output to the MUX unit 50 together with the character data. In the MUX unit 50, the character data, the image packet, the voice packet, and the program information extracted from the information storage unit 40 according to an instruction from the control unit 20 are combined into a TS and output. This keyword is a key for searching, and may be a specific voice signal, for example, a voice signal of a specific singer or a specific music melody, and can be set and selected as appropriate.
[0027]
Thus, since the contents of the broadcast program are stored as character data in the information storage unit 40, when a desired keyword is input, the control unit 20 searches the character data stored by the keyword, and if they match, the desired content is stored. Is played. Further, the EPG can be compared and analyzed with character data to display an outline of program contents as character data, or the character data can be displayed as it is or after conversion such as translation.
[0028]
Note that the voice recognition unit 32 may recognize a unit of arbitrary length, and the time information added by the synthesis unit 34 may be added to character data of an identification unit of arbitrary length instead of every word. Further, if a plurality of pieces of character data of the identification unit are combined and time information is added to reduce the amount of time information, the capacity of the storage medium can be saved. Furthermore, sounds other than natural sounds and sounds such as musical instruments are separately distinguished, and character data that can be arbitrarily identified can be assigned and processed.
[0029]
Next, a program recording apparatus according to the second embodiment will be described with reference to FIG. 1 referred to in the first embodiment.
[0030]
The program recording apparatus according to the second embodiment converts the appearance frequency of character data that has been voice-recognized into data, and since the apparatus configuration is the same as that of the first embodiment, description thereof is omitted. Also, when the same operation is performed in each part, the description is omitted.
[0031]
The speech recognition unit 32 of the speech processing unit 30 in FIG. 1 recognizes the speech signal restored by the DEC unit 31 and converts it into character data. Further, the appearance frequency of each word A to D. ... are examined and converted into data as shown in the histogram of FIG. 5, and the appearance frequencies a to c are added to the words A to D to generate Hist data as shown in FIG. . The Hist data is sent to the information storage unit 40 together with the character data, and is recorded and stored in the recording medium 41.
[0032]
When a keyword is input by the user, the keyword and the Hist data are referred to, and a list of broadcast program names having the highest appearance frequency of the keyword is sequentially displayed from the recorded broadcast programs, or further combined with the EPG. It is also possible to search for broadcast programs to be recorded. Of course, the appearance frequency may be the appearance frequency of character data of an arbitrary length instead of a word.
[0033]
【The invention's effect】
As described above, when a program is recorded, the audio signal of the program is recognized and stored as character data, so that the user can search for a program recorded with an arbitrary keyword and one scene of the program . This enables more efficient and detailed search.
[0034]
Furthermore, an outline of program contents can be obtained by using EPG or the like. In addition, it becomes easy to search for a program that does not have an EPG, and in addition, it can be used as subtitles by displaying character data as it is or after being translated.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a main part of a program recording apparatus according to the present invention.
FIG. 2 is an explanatory diagram showing a time series when a voice packet is recognized by voice and converted into character data in the program recording apparatus according to the present invention.
FIG. 3 is a conceptual diagram illustrating an example of a format format of character data according to the present invention.
FIG. 4 is a conceptual diagram showing a file structure of data recorded on a recording medium of the program recording apparatus according to the present invention.
FIG. 5 is a conceptual diagram showing a state in which the appearance frequency of character data is made into a histogram in the program recording apparatus according to the present invention.
FIG. 6 is a conceptual diagram illustrating an example of a format format of character data to which appearance frequency is added.
[Explanation of symbols]
10; DMUX section, 20; control section, 30; voice processing section, 31; DEC section, 32; voice recognition section, 33; shift section, 34; synthesis section, 40; information storage section, 41; MUX section

Claims (7)

Record the program with voice recognition,
The voice recognition result is converted into text data, recorded and stored,
Analyzing the content of the program program from the text data, recording and storing the content of the analyzed program program,
A program recording method for searching stored text data using a specific audio signal as a key ,
A program recording method, wherein the appearance frequency of each voice recognition unit is recorded together with the result of the voice recognition.   Means for recording a program while performing speech recognition;
  Means for converting the voice recognition result into text data and recording and storing the result;
  Means for analyzing the contents of the program program from the text data, recording and storing the contents of the analyzed program program;
  Means for searching for stored text data using a specific audio signal as a key, and a program recording device comprising:
  A program recording apparatus comprising means for recording the appearance frequency of each voice recognition unit together with the result of the voice recognition.   Means for recording a program while performing speech recognition;
  Means for converting the voice recognition result into text data and recording and storing the result;
  A program recording / playback apparatus comprising: a playback unit that compares an input keyword with the stored text data and plays back a program related to the matched text data,
  A program recording / reproducing apparatus comprising means for recording the appearance frequency of each voice recognition unit together with the voice recognition result.   In the program recording / reproducing apparatus according to claim 3,
  A program recording / reproducing apparatus, wherein the voice recognition units are displayed in the descending order of appearance frequency, are appropriately selected, and a program related to the text data of the selected voice recognition units is reproduced.   In the program recording / reproducing apparatus according to claim 3,
  A program recording / reproducing apparatus comprising means for analyzing the contents of a program program from the text data, and recording and storing the contents of the analyzed program.   In the program recording / reproducing apparatus according to claim 5,
  A program recording / reproducing apparatus, characterized in that the contents of the stored program programs are displayed and appropriately selected, and a program related to the contents of the selected program program is reproduced.   Record the program with voice recognition,
The voice recognition result is converted into text data, recorded and stored,
A program recording / playback method for playing back a program related to matched text data by comparing an input keyword and the stored text data,
  A program recording / reproducing method, wherein the appearance frequency of each voice recognition unit is recorded together with the result of the voice recognition.

Images