JP4148261B2 - Karaoke equipment - Google Patents

Description

  The present invention relates to karaoke singing in a karaoke apparatus, and more particularly to speech recognition technology for lyrics sung in karaoke.

  Conventionally, a scoring function is known as an incidental function of a karaoke apparatus. This scoring function is the following function. That is, singing data indicating the singing state such as pitch, volume or tempo uttered by the singer is generated by sampling the voice signal of the singer input from the microphone. Next, the singing data is compared with scoring reference data such as main melody part data in karaoke data, and a predetermined score is given based on the comparison result to generate scoring data. When the singing part ends, the scores in the scoring data are totaled to calculate a total score. The total score is displayed as it is on a scoreboard or a display, or a video reflecting the total score such as a video including a predetermined message or predetermined expression content is output to the display (for example, see Patent Document 1).

  In addition, when multiple people sing, such as duet songs, men and women based on the input frequency of audio signals from a karaoke device (for example, refer to Patent Document 2) having a scoring function for scoring each person's singing voice or two microphones. There is a karaoke apparatus (see, for example, Patent Document 3) such as a singing scoring system that determines the score and scores separately.

  In the karaoke apparatus having the scoring function as described above, the karaoke singing ability is scored based on the singing data indicating the singing state such as the pitch, volume or tempo uttered by the singer.

However, in karaoke singing, there has been no karaoke apparatus that recognizes the lyric lyrics.
If not limited to the technical field of the karaoke device, such as a noise removal device that can sufficiently recognize speech even in an environment where noise, output sound of an audio device, etc. exist in addition to a voice command for control. There is a voice recognition device (see, for example, Patent Document 4).
Japanese Patent No. 3261990 JP 2000-330580 A Japanese Patent Laid-Open No. 11-282478 Japanese Patent Laid-Open No. 5-73091

  However, in recent years, games have emerged that use karaoke devices to compete for singing karaoke songs without making mistakes. In this game, judges and others judged whether or not they sang karaoke songs without making mistakes. In other words, through the ears of judges, etc., listen to the lyrics sung by karaoke, recognize the lyrics, compare with the lyrics of the karaoke songs that were sung karaoke, and sing the karaoke song without mistakes. I was judging.

  Therefore, this game cannot be enjoyed when one person sings a karaoke using a karaoke apparatus. Also, even when singing karaoke with a large number of people, for example, one user listens to the lyrics sung by karaoke, recognizes the lyrics, and compares the lyrics of the karaoke performance karaoke song with the wrong karaoke song lyrics. For companions who decide whether or not to sing without karaoke, they could not fully enjoy karaoke performance.

  There is also a game that decides in advance the lyrics that should not be sung from the lyrics of karaoke songs (hereinafter referred to as NG words) and competes whether or not the NG words were sung during the karaoke song. It was. Also in this game, judges and the like judged whether or not they sang NG words during karaoke singing.

These games are required to be automated without relying on a person such as a judge as in the case of “singing a song”.
However, when recognizing karaoke-sung lyrics, the karaoke performance sound is emitted from the karaoke device, so the karaoke performance sound is also input to the microphone together with the sound of singing the karaoke song. Therefore, there has been a problem that the karaoke performance sound itself other than the voice for singing karaoke becomes noise and the recognition rate of voice recognition is lowered.

  As a conventional technique for recognizing voices in a place with a large ambient noise, there is a technique for extracting and deleting ambient sounds such as car engine sounds and indoor duct sounds. Note that such ambient sounds are noises having a constant volume and frequency and have a short period, so that the time until the ambient sounds are extracted (hereinafter also referred to as a time lag) is short. Therefore, even if there is a time lag, ambient sounds can be extracted after the time lag has elapsed.

  However, in the case of a karaoke device, one cycle of karaoke performance sound is one part of one piece of music, so the cycle is very long compared to ambient sounds such as the engine sound of the car and the duct sound of the room. . Moreover, since the frequency of the karaoke performance sound and the volume of the music performance (for example, fortesimo, pianissimo, etc.) change irregularly, it takes a long time to extract the karaoke performance sound as the ambient sound. Therefore, it is difficult to erase the karaoke performance sound as the ambient sound. Also, unlike the voice of daily conversation, the voice of karaoke singing is pronounced to match the karaoke performance sound (ambient sound), so the karaoke singing lyrics are distinguished from the karaoke performance sound (ambient sound). It was difficult to recognize.

The karaoke apparatus of the present invention made to solve the above-mentioned problems (1: In this section, in order to facilitate the understanding of the invention, “the best mode for carrying out the invention” is necessary as necessary. The components described in the column are shown in parentheses, but this description does not mean that the scope of the claims is limited. The voice signal input means (M2) for inputting the voice signal of karaoke singing ), Music data storage means (M10) for storing the music data of the karaoke music, and the music data stored in the music data storage means is reproduced as a sound signal, and the reproduced sound signal and the audio signal input means are input. The karaoke performance reproducing means (M14) for outputting the karaoke singing voice signal to the speaker (M20) and the sound signal output from the karaoke performance reproducing means. The first signal to be compared with the second signal corresponding to the voice signal of the karaoke song input from the voice signal input means, and a voice recognition signal is generated by subtracting the first signal from the second signal. Based on the first generation means (M70) and the speech recognition signal generated by the first generation means, a gain for the first generation means to subtract the first signal from the second signal is set. Output from the first gain setting means (M72), the voice recognition means (M76) for recognizing the lyrics of the karaoke song based on the voice recognition signal generated by the first generation means, and the karaoke performance reproduction means. The first signal corresponding to the received sound signal is compared with the second signal corresponding to the voice signal of the karaoke song input from the voice signal input means, and the first signal is subtracted from the second signal. Second generation to generate the signal Based on the stage (M30), the pitch extraction means (M6) for extracting pitch data from the scoring signal generated by the second generating means, and the scoring signal generated by the second generating means, Second gain setting means (M32) for setting a gain for the second generation means to subtract the first signal from the second signal, and the pitch of the karaoke song melody stored in the song data storage means A singing scoring means for scoring karaoke singing for each predetermined section based on the calculated pitch difference and calculating the pitch difference between the data and the pitch data of the scoring signal extracted by the pitch extracting means ( M24). Then, in the case of the prelude period during the karaoke song performance period or the period in which the interlude period is not sung, the first gain setting means sets the gain so that the signal level of the speech recognition signal is minimized, and the second gain The setting means sets the gain so that the signal level of the scoring signal is minimized.

  As described above, the karaoke apparatus of the present invention uses the voice recognition signal generated by subtracting the sound signal corresponding to the karaoke performance sound emitted from the speaker even when the karaoke performance sound is emitted from the speaker. Since the lyrics of the karaoke song are recognized based on this, an appropriate speech recognition result can be obtained. Specifically, the first signal corresponding to the sound signal output from the karaoke performance reproducing means is a performance signal of the karaoke performance sound, and the karaoke performance sound is emitted from the speaker. Therefore, in the case of voice recognition, the karaoke performance sound emitted from the speaker is input to the microphone separately from the voice of the singer. That is, the karaoke performance sound emitted from the speaker is input to the second signal corresponding to the karaoke singing voice signal input from the voice signal input means. However, the first generation means (M70) compares the first signal with the second signal and generates a voice recognition signal obtained by subtracting the first signal from the second signal, so that it is released from the speaker. The sound signal corresponding to the sung karaoke performance sound is reduced. And since a voice recognition means (M76) recognizes the lyrics of a karaoke song based on the signal for voice recognition produced | generated by the 1st production | generation means, it can obtain a suitable voice recognition result.

  In addition, as described in claim 2, the lyrics of the karaoke song may be scored based on the lyrics data of the karaoke song recognized by the voice recognition means of the karaoke apparatus according to claim 1.

  That is, in the karaoke apparatus according to claim 1, the music data storage means stores music data including lyrics data of the karaoke music, and the lyrics data and music data storage means of the karaoke song recognized by the voice recognition means. The lyric scoring means for comparing the lyric data of the karaoke song, extracting the quantity of the different lyric data, and scoring the lyrics of the karaoke song based on the extracted quantity and the lyric data quantity of the karaoke song (12) may be provided.

  According to the karaoke apparatus configured in this way, the lyrics data of the karaoke song recognized by the voice recognition means is compared with the lyrics data of the karaoke song stored in the song data storage means, and the lyrics of the karaoke song are scored. can do. Specifically, the lyrics data of the karaoke song recognized by the voice recognition means, for example, a plurality of words sung by the karaoke song and a plurality of words on the lyrics data of the karaoke song stored in the song data storage means are compared. To do. Then, the number of different lyrics data, for example, the number of different words from a plurality of words is extracted, and the lyrics of the karaoke song based on the extracted number of words and the number of words included in the lyrics data of the karaoke song Scoring. More specifically, for example, if there are five different words and the number of words included in the lyrics data of the karaoke song is 100, (100-5) /100=0.95. , You can score lyrics with 95 points out of 100.

  Furthermore, in order to improve the game performance of lyric scoring, as described in claim 3, karaoke performance may be stopped if the lyrics of the karaoke song recognized by the voice recognition means of the karaoke apparatus are erroneous by a predetermined number or more. It is done.

  That is, in the karaoke apparatus according to claim 1, the music data storage means stores music data including lyrics data of the karaoke music, and the lyrics data and music data storage means of the karaoke song recognized by the voice recognition means. Comparing the lyrics data of the karaoke song and extracting the quantity of the different lyric data. If it is determined that the extracted quantity is a predetermined number or more, the karaoke performance playback means is controlled to The control means (12) which stops reproduction | regeneration of this is provided.

  According to the karaoke apparatus configured in this way, even when the karaoke song is being performed, if a predetermined number of lyrics are sung in error, the karaoke performance is stopped, so that the game performance can be improved. Note that the predetermined number is set according to the level of each karaoke singer, for example, “1” for a skilled karaoke singer, “3” for an intermediate singer, or “5” for a beginner. It is a numerical value.

  Further, as described in claim 4, when NG words that should not be sung are displayed in advance from the lyrics of karaoke singing and NG words are sung during karaoke singing, It is also possible to cancel.

That is, the karaoke apparatus according to claim 1 includes display means (M26) capable of displaying lyrics data of karaoke songs.
The music data storage means stores music data including lyrics data of karaoke music.

  Then, specific lyric data is extracted from the music data stored in the music data storage means, the display means is controlled to display the extracted lyric data, and the karaoke song recognized by the voice recognition means Control means (12) is provided for controlling the karaoke performance reproduction means to stop the reproduction of the karaoke performance when the lyrics data is compared with the extracted lyrics data and it is determined that there is the same lyrics data.

According to the karaoke apparatus configured in this way, it is possible to enjoy a game of competing whether or not an NG word has been sung during karaoke singing.
In addition, in the karaoke apparatus according to claim 2, as a condition for scoring lyrics, the singing scoring result may be limited to a predetermined value or more as described in claim 5.

That is, the karaoke apparatus according to claim 2 further includes voice recognition storage means (M10 ) for storing lyrics data of the karaoke song recognized by the voice recognition means.

  Then, the lyrics scoring means determines whether or not the singing score result of the karaoke singing by the singing scoring means is greater than or equal to a predetermined value for each predetermined section. The karaoke song lyrics data is read out, the karaoke song lyrics data read out is compared with the karaoke song lyrics data stored in the song data storage means, and the quantity of the different lyric data is extracted and extracted. The karaoke singing lyrics may be scored based on the number of karaoke songs and the number of lyrics data of the karaoke songs.

  According to the karaoke apparatus configured in this way, as a condition for scoring the lyrics, the singing scoring result is limited to a predetermined value or more, for example, 80 points or more, so that the recognition rate of voice recognition can be prevented from being lowered. That is, for example, when the timing is too shifted, the volume level is low, or the pitch level is high, and the singing score is not more than a predetermined value, for example, 80 points or more, it is considered that the recognition rate of voice recognition also decreases. Because.

  Also, in the karaoke apparatus according to claim 3, as a condition for playing a game in which karaoke performance is stopped if the lyrics are incorrect, the singing scoring result is limited to a predetermined value or more as described in claim 6. Good.

That is, the karaoke apparatus according to claim 3 further includes voice recognition storage means (M10 ) for storing lyrics data of the karaoke song recognized by the voice recognition means.

  Then, the control means determines whether or not the singing result of the karaoke singing by the singing grading means is greater than or equal to a predetermined value for each predetermined section, and if it is determined that it is greater than or equal to the predetermined value, it is stored in the voice recognition storage means. The karaoke song lyrics data is read out, the karaoke song lyrics data read out is compared with the karaoke song lyrics data stored in the song data storage means, and the quantity of the different lyric data is extracted and extracted. If it is determined that there is a predetermined number or more, the karaoke performance playback means may be controlled to stop playback of the karaoke performance.

  According to the karaoke apparatus configured in this way, as a condition for playing a game in which karaoke performance is stopped if the lyrics are incorrect, the singing scoring result is limited to a predetermined value or more, for example, 80 points or more. It is possible to prevent a decrease in the recognition rate of voice recognition in a high game.

  Further, in the karaoke apparatus according to claim 4, as described in claim 7, the singing scoring result is a condition for playing a game for competing whether or not the NG word was sung during karaoke singing. It should be limited to a predetermined value or more.

That is, the karaoke apparatus according to claim 4 further includes voice recognition storage means (M10 ) for storing lyrics data of the karaoke song recognized by the voice recognition means.

  Then, the control means extracts specific lyric data from the music data stored in the music data storage means, controls the display means to display the extracted lyric data, and singing a score for each predetermined section It is determined whether or not the singing score result of the karaoke song by the means is greater than or equal to a predetermined value, and if it is determined that it is greater than or equal to the predetermined value, the lyrics data of the karaoke song stored in the voice recognition storage means is read and read If the lyrics data of the karaoke song is compared with the extracted lyrics data and it is determined that there is the same lyrics data, the playback of the karaoke performance may be stopped by controlling the karaoke performance playback means.

  According to the karaoke apparatus configured in this manner, the singing scoring result is limited to a predetermined value or more, for example, 80 points or more as a condition for playing a game for competing whether or not the NG word has been sung during karaoke singing. Thus, it is possible to prevent a decrease in the recognition rate of voice recognition in a game with higher competitiveness.

Embodiments of the present invention will be described below with reference to the drawings.
[Description of schematic configuration of karaoke equipment function]
FIG. 1 is a diagram showing a schematic configuration centering on the function of the karaoke apparatus. Among the functions shown in FIG. 1, a data extraction unit M6, a singing comparison unit M8, a sequencer M14, a scoring unit M24, a singing difference extraction unit M34, a switch control unit M36, a singing adjustment unit M37, a voice recognition unit. About M76, the difference extraction part M74 for lyrics, the adjustment part M77 for lyrics, and the determination part M80, it implement | achieved mainly with CPU and software.

  First, the karaoke apparatus has a scoring function, and the singing accuracy is determined by digitizing and capturing the voice signal of the karaoke song and comparing it with the pitch of the guide melody, which is a song melody.

Moreover, the karaoke apparatus has a voice recognition function, digitizes the voice signal of the karaoke singing, and recognizes the lyrics.
Next, as shown in FIG. 1, a microphone M2 (hereinafter, abbreviated as a microphone) for a singer to input karaoke singing voice is connected to an amplifier M18 and to an A / D converter M4. ing. In addition, music data including karaoke performance data, guide melody data, and lyrics data is stored in the data storage unit M10. Of the song data, one song selected by the karaoke singer is read to the execution memory M12, and is sequentially read by the sequencer M14 during performance. The karaoke performance data read out by the sequencer M14 is input to the musical tone generator M16. The guide melody data sequentially read by the sequencer M14 is input to the singing comparison unit M8. The musical sound generator M16 generates a performance signal for karaoke music based on the input karaoke performance data, and this performance signal is input to the amplifier M18. The amplifier M18 amplifies the karaoke performance signal and the singing voice signal input from the microphone M2, and outputs the amplified signal to the speaker M20.

  Since this guide melody data corresponds to the singing melody of karaoke music, it is also used when realizing a function output from the speaker M20 together with the accompaniment music as a so-called guide melody function. The guide melody function can be switched on / off by the user (karaoke singer), and the user who thinks that the function is necessary can be turned on by operating an operation panel or a remote control (not shown). Then, not only the karaoke accompaniment but also the singing melody is output from the speaker M20 as a guide melody, and it is possible to sing with reference to it. On the other hand, if the function is turned off, the guide melody is not output from the speaker M20, and the guide melody data is used only for scoring.

  In the following description, it is assumed that the guide melody function is on. Therefore, the karaoke song performance signal and the guide melody signal are added to the singing voice signal input from the microphone M2. In the following description, the term “singing voice signal” refers to a singing voice signal in which the performance signal of the karaoke song and the sound signal of the guide melody are added, and the term “singing voice” simply refers to the performance of the karaoke song. It refers to a singing voice to which a sound and a guide melody are added.

  On the other hand, the singing voice signal input to the A / D converter (ADC) M4 is converted into a digital signal and then input to the singing signal generation unit M30 and the lyrics signal generation unit M70. The musical tone generator M16 described above generates a karaoke song performance signal and a guide melody sound signal based on the inputted karaoke performance data and the guide melody data. The karaoke song performance signal and the guide melody sound are generated. The signal is input to the singing variable gain amplifier M32 and the lyrics variable gain amplifier M72.

  The performance signal of the karaoke song and the sound signal of the guide melody input to the singing variable gain amplifier M32 are amplified by the singing variable gain amplifier M32 and input to the singing signal generation unit M30. The singing signal generation unit M30 is a singing voice signal converted into a digital signal input from an A / D converter (ADC) M4, a karaoke song performance signal and a guide melody sound signal input from a singing variable gain amplifier M32. A scoring signal is generated from the data and input to the data extraction unit M6 and to the singing difference extraction unit M34. The scoring signal input to the singing difference extraction unit M34 is switched by a switch control unit M36 via a switch M38 that is controlled to be connected (hereinafter also simply referred to as “on”) or disconnected (hereinafter also simply referred to as “off”). To the singing adjustment unit M37. Then, the singing adjustment unit M37 instructs the singing variable gain amplifier M32 to gain the signal level of the scoring signal input to the singing adjustment unit M37.

  The karaoke song performance signal and the guide melody sound signal input to the lyrics variable gain amplifier M72 are amplified by the lyrics variable gain amplifier M72 and input to the lyrics signal generation unit M70. The lyric signal generator M70 is a singing voice signal converted into a digital signal input from an A / D converter (ADC) M4, a karaoke song performance signal and a guide melody sound signal input from a lyric variable gain amplifier M72. A speech recognition signal is generated from the above and input to the speech recognition unit M76 and to the lyrics difference extraction unit M74. The voice recognition signal input to the lyric difference extraction unit M74 has a switch M78 that is controlled by the switch control unit M36 to be connected (hereinafter simply referred to as ON) or disconnected (hereinafter also simply referred to as OFF). To the lyrics adjustment unit M77. Then, the lyrics adjustment unit M77 instructs the lyrics variable gain amplifier M72 to gain such that the signal level of the speech recognition signal input to the lyrics adjustment unit M77 is minimized. Also, the karaoke performance data read by the sequencer M14 is input to the musical sound generating unit M16 via the switch M40, which is controlled by the switch control unit M36 to be connected or disconnected.

  Here, the switch control unit M36 determines whether or not it is a singing period from the music data read by the sequencer M14, and controls the switch M38, the switch M40, and the switch M78.

  2 shows on / off changes of the switch M40, the switch M38 and the switch M78 with respect to the elapsed time after the end of the karaoke performance, the karaoke performance period, the singing variable gain amplifier M32 and the lyrics variable gain amplifier. It is explanatory drawing which shows the change of the output signal level from M72. Before the karaoke performance, the switch M40, the switch M38 and the switch M78 are turned off. However, when the sequencer M14 reads the music data and enters the karaoke performance period, the switch control unit M36 controls the switch M40 to be turned on. It is kept on until the end of the karaoke performance, and is turned off when the karaoke performance ends. When the switch control unit M36 determines whether or not it is a singing period from the music data read out by the sequencer M14 and determines that it is not a singing period, that is, when it determines that it is a prelude period or an interlude period, M38 and switch M78 are controlled to be turned on, and when it is determined that it is a singing period, switch M38 and switch M78 are controlled to be turned off. Here, “the switch control unit M36 determines whether or not it is a singing period from the music data read out by the sequencer M14” is described later in [Description of NG word game processing of voice control unit 24]. As described above, “whether it is a period in which the prelude period or the interlude period is not sung by the section division data indicating the start point of the singing section of the karaoke song stored in the control data track delivered from the sequence program for performing the karaoke performance. To judge.

  When it is determined that it is not a singing period, that is, when it is determined that it is a prelude period or an interlude period, the switch M38 is controlled to be turned on, and is input from the singing difference extraction unit M34 to the singing adjustment unit M37. The singing adjustment unit M37 instructs the singing variable gain amplifier M32 so as to minimize the signal level of the scoring signal (see FIG. 2). Therefore, when it is determined from the music data that it is not the singing period, that is, during the prelude period or the interlude period, the signal level of the scoring signal is adjusted to be minimum.

  On the other hand, if it is determined that it is the singing period, the switch M38 is controlled to be turned off, so that the singing variable gain amplifier M32 is not instructed by the singing adjustment unit M37, and the singing variable gain amplifier The output signal level from M32 is fixed (see FIG. 2). Then, the data extraction unit M6 to which the scoring signal generated by the singing signal generation unit M30 is input calculates the pitch from the digitized scoring signal every 50 ms, for example, and the singing comparison unit is used as pitch data. Input to M8. The singing comparison unit M8 compares the pitch data of the scoring signal input from the data extraction unit M6 with the pitch data of the guide melody (hereinafter also simply referred to as reference) input from the sequencer M14 in real time. Note that 50 ms corresponds to a thirty-second note with a metronome tempo of 120, and has a sufficient resolution for extracting the characteristics of singing.

  The singing comparison unit M8 calculates the difference between the scoring signal and the reference every 50 ms in accordance with the timing at which the scoring signal is input. This is output to the scoring unit M24 as real-time difference data (pitch difference data). In the scoring unit M24, the pitch difference data is totaled for each section of the karaoke music and stored in the data storage unit M10. And in the scoring part M24, the difference data of each section is totaled and a scoring result is obtained.

  Further, when it is determined that it is not a singing period, that is, when it is determined that it is a prelude period or an interlude period, the switch M78 is controlled to be turned on and input from the lyrics difference extracting unit M74 to the lyrics adjusting unit M77. Then, the lyrics adjusting unit M77 instructs the lyrics variable gain amplifier M72 so as to minimize the signal level of the voice recognition signal (see FIG. 2). Therefore, when it is determined from the music data that it is not the singing period, that is, during the prelude period or the interlude period, the signal level of the voice recognition signal is adjusted to be minimum.

  On the other hand, when it is determined that it is the singing period, the switch M78 is controlled to be turned off, so that the lyrics variable gain amplifier M72 is not instructed by the lyrics adjustment unit M77 to have a gain and the lyrics variable gain amplifier. The output signal level from M72 is fixed (see FIG. 2). Then, the voice recognition unit M76, to which the speech recognition signal generated by the lyrics signal generation unit M70 is input, recognizes the lyrics and stores them in the data storage unit M10 as lyrics data. The determination unit M80 extracts specific lyric data of the selected music from the music data stored in the data storage unit M10 and displays it on the display unit M26, and the scoring stored in the data storage unit M10 for each predetermined section. When it is determined that the result is a predetermined value, for example, 80 points or more, the speech-recognized lyrics data stored in the data storage unit M10 is read, and the speech-recognized lyrics data and the extracted lyrics are read When it is determined that there is the same lyric data by comparing with the data, the sequencer M14 is controlled to stop the karaoke performance.

  In FIG. 1 showing a schematic configuration centering on the function of the karaoke apparatus, the microphone M2 corresponds to “speech signal input means”, and the data storage unit M10 is “music data storage means” and “voice recognition storage means”. It corresponds to. The sequencer M14 corresponds to “karaoke performance reproduction means”, the singing signal generation unit M30 corresponds to “second generation means”, and the data extraction unit M6 corresponds to “pitch extraction means”. The switch control unit M36, the switch M38, the switch M40, the singing difference extraction unit M34, the variable gain amplifier M32, and the singing adjustment unit M37 correspond to the “second gain setting unit”. The scoring unit M24 corresponds to “singing scoring means”. The speaker M20 corresponds to a “speaker”, and the display unit M26 corresponds to a “display unit”.

  The lyrics signal generation unit M70 corresponds to the “first generation unit”, and the voice recognition unit M76 corresponds to the “voice recognition unit”. The switch control unit M36, the switch M78, the switch M40, the lyrics difference extracting unit M74, the lyrics variable gain amplifier M72, and the lyrics adjusting unit M77 correspond to the “first gain setting means”. To do. The determination unit M80 corresponds to a “control unit”.

  Next, the scoring signal and reference will be described with reference to FIG. The dotted line illustrated in FIG. 3 is obtained by converting the guide melody that is a reference into pitch data, and general guide melody data is mechanically very accurate. On the other hand, the solid line illustrated in FIG. 3 shows the scoring signal converted to pitch data. The pitch data of the scoring signal varies up and down from the value indicated by the pitch data of the guide melody, and there is no pitch data of the scoring signal in the prelude and interlude.

Although FIG. 1 shows a schematic configuration centering on the function of the karaoke apparatus, FIG. 4 is a block diagram showing a specific hardware configuration of the karaoke apparatus.
[Description of configuration of karaoke apparatus 1]
FIG. 4 is a block diagram showing a configuration of the karaoke apparatus 1. As shown in FIG. 4, the karaoke apparatus 1 includes a control unit 12 that controls the operation of the karaoke apparatus 1 as a whole, an interface unit 14 for connecting the karaoke apparatus 1 to the network 100, music that shows the accompaniment content and lyrics of a performance tune. A hard disk (HDD) 16 for storing data, video data, etc., an operation unit 18 comprising a plurality of keys and switches, an infrared communication unit 20 for receiving infrared signals from the remote control terminal 2 and mobile phone by infrared communication, an operation unit An operation processing unit 22 that processes a signal from 18, an audio signal (signal related to sound and sound) of a performance music is generated from music data stored in the hard disk 16, and the generated audio signal and audio input from the microphone 25 are generated. A voice control unit 24 that amplifies the signal and outputs the amplified signal to the speaker 28; IDI sound source 30, video RAM 32 for temporarily storing video information, video playback unit 34 for controlling playback of video based on video data, video information stored in video RAM 32 and display of video played back by video playback unit 34 A video control unit 38 for controlling display in the unit 36 is provided.

  The control unit 12, interface unit 14, HDD 16, infrared communication unit 20, operation processing unit 22, video RAM 32, video playback unit 34, and video control unit 38 are connected by a bus 39. The control unit 12 and the audio control unit 24 are connected by a USB 40. In addition, the control part 12 and the audio | voice control part 24 perform the various processes mentioned later.

  Among these, as illustrated in FIG. 5A, the HDD 16 stores a song data memory area 50 for storing song data, a word data memory area 52 for storing word data extracted from the song data, and pitch difference data. There is provided a scoring information memory area 54 for storing scoring information according to. The music data stored in the music data memory area 50 includes word data used in header information, MIDI data, title data, and lyrics, as illustrated in FIG. The header information has a MIDI data size, a title data size, and a word data size. The data structure of the word data used in the lyrics includes the number of registered words, the first word size, the first word, the second word size, the second word, the nth word size, and the nth word. The data structure continues to the word. For example, as shown in FIG. 6B, data indicating “3 words” corresponding to the registered word count, data indicating “8 bytes” corresponding to the first word size, Corresponding data indicating “Idoru”, data indicating “6 bytes” corresponding to the second word size, data indicating “nice” corresponding to the second word, and “10 bytes” corresponding to the third word size The data structure continues to “I love you” corresponding to the third word.

  The MIDI data included in the music data has a music track, a guide melody track, and a control data track as illustrated in FIG. 5B. In the music track, a track of various parts including a melody track and a rhythm track is formed. The guide melody track stores the melody sequence data of the karaoke song, that is, the melody that the singer should sing. The control data track stores section division data indicating the starting point of the singing section of the karaoke song.

  The scoring information corresponding to the pitch difference data stored in the scoring information memory area 54 has a data structure in which the pitch difference data is associated with the scoring as illustrated in FIG. For example, “100 semitones” as pitch difference data is associated with “100 points” as a scoring.

  Further, in the ROM (not shown) built in the voice control unit 24, as illustrated in FIG. 5C, the lyrics recording area 60 for the karaoke song, the guide melody buffer 62, the reference data register 64, and the difference data storage area. 66 is provided. In the lyrics recording area 60 of the karaoke song, the speech-recognized lyrics are stored. The guide melody buffer 62 temporarily stores the read guide melody data. The reference data register 64 stores a reference extracted from the guide melody data (that is, pitch data of the guide melody). The difference data storage area 66 stores difference data between the reference and the singing voice. The reference data register 64 is a pitch data register, and the difference data storage area 66 is a pitch difference data storage area.

  In the present embodiment, the microphone 25 corresponds to “audio signal input means”, the HDD 16 corresponds to “music data storage means”, and the ROM built in the audio control unit 24 corresponds to “voice recognition storage means”. To do. Further, the voice control unit 24 performs “karaoke performance reproduction means”, “first generation means”, “second generation means”, “pitch extraction means”, “first gain setting means”, and “second gain setting means”. It corresponds to “gain setting means”, and the control unit 12 corresponds to “control means” and “singing scoring means”. The speaker 28 corresponds to “speaker”, the voice recognition unit 10 corresponds to “voice recognition unit”, and the display unit 36 corresponds to “display unit”.

[Description of NG word game process of control unit 12]
Below, the procedure of "NG word game process of the control part 12" which the control part 12 of the karaoke apparatus 1 performs is demonstrated based on the flowchart of FIG. 7, FIG.

  In the following description, it is assumed that a karaoke song is selected by a user (karaoke singer). Specifically, the music selection number data of the karaoke song received by the operation unit 18 is transmitted to the control unit 12 by the operation processing unit 22, but the control unit 12 receives the music selection number data and receives the music selection number. It is assumed that the number data is being transmitted to the voice control unit 24.

  The game start instruction data received by the operation unit 18 is transmitted to the control unit 12 by the operation processing unit 22, but the control unit 12 determines whether or not the game start instruction data has been received (S110). . When the game start instruction data transmitted from the operation processing unit 22 is received (S110: YES), the music data corresponding to the music selection number of the selected karaoke music is provided in the HDD 16 (a) in FIG. Is read from the music data memory area 50 illustrated in FIG. Then, the title data and the word data used in the lyrics exemplified in FIG. 6A are read from the read music data, and the read word data are stored in the word data memory area 52 (S114). . When the process of S114 is completed, the process of S116 is executed.

  In the process of S116, the title read in the process of S114 is displayed on the display unit 36. Specifically, as illustrated in FIG. 16A, the video control unit 38 is controlled so that “red girls” and the like corresponding to the title read in the process of S <b> 114 are displayed on the display unit 36. . Next, the NG word is displayed on the display unit 36, and the NG word data is transmitted to the voice control unit 24 (S118). Specifically, NG word data is selected from the word data stored in the word data memory area 52 in the process of S114, and “I love you” corresponding to the selected NG word as illustrated in FIG. 16B. The video control unit 38 is controlled so as to be displayed on the display unit 36. Then, the NG word data is transmitted to the voice control unit 24. When the process of S118 ends, the process of S120 is executed.

  In the process of S120, a karaoke performance start signal is transmitted to the voice control unit 24. Then, a voice recognition start signal is transmitted to the voice control unit 24 (S122). Furthermore, a section singing start signal is transmitted to the voice control unit 24 (S124, see FIG. 8). When the process of S124 is completed, the process of S126 is executed.

  The comparison result data is transmitted from the voice control unit 24 to the control unit 12 (this transmission process will be described later). The control unit 12 determines whether or not the comparison result data has been received. (S126). When the comparison result is received from the voice control unit 24 (S126: YES), the process of S128 is executed. On the other hand, when the comparison result is not received from the voice control unit 24 (S126: NO), the process of S138 is executed.

  In the process of S128, it is determined whether or not an NG word has been sung. For example, it is assumed that if the number of NG word data in the lyrics data of karaoke singing is one or more, the NG word is sung. If it is determined that the NG word has been sung (S128: YES), a karaoke performance stop signal is transmitted to the voice control unit 24 (S130). When the process of S130 is completed, the process of S132 is executed. On the other hand, when it is determined that the NG word is not sung (S128: NO), the process of S138 is executed.

In the process of S132, a voice recognition end signal is transmitted to the voice control unit 24.
The data of the NG word lyrics screen is transmitted from the voice control unit 24 to the control unit 12 (this transmission process will be described later), but the control unit 12 determines whether or not the data of the NG word lyrics has been received. Is determined (S134). When NG word lyrics data is received (S134: YES), an NG word lyrics screen is displayed on the display unit 36 (S136). Specifically, as illustrated in FIG. 16 (c), “I saw me” or the like corresponding to the lyrics immediately before singing the NG word is displayed on the display unit 36, and FIG. 16 (e). The video control unit 38 is controlled to display “Still! Sorry” on the display unit 36 as illustrated in FIG.

Then, when the process of S136 ends, the present “scoring process of the control unit 12” ends.
In the process of S138, it is determined whether or not a karaoke performance end signal has been received from the voice control unit 24. If no karaoke performance end signal is received from the voice control unit 24 (S138: NO), the process returns to S124, and the above-described process is executed. On the other hand, when a karaoke performance end signal is received from the voice control unit 24 (S138: YES)
A voice recognition end signal is transmitted to the voice control unit 24 (S140). Then, the completion result is displayed on the display unit 36 (S142). Specifically, as illustrated in FIG. 16D, the video control unit 38 is controlled so that “Done! Congratulations” is displayed on the display unit 36.

Then, when the process of S142 is completed, the present “scoring process of the control unit 12” ends.
[Description of NG word game process of voice control unit 24]
Next, the procedure of “NG word game processing of the voice control unit 24” executed by the voice control unit 24 of the karaoke apparatus 1 will be described with reference to the flowcharts of FIGS. The operation program related to the NG word game process is executed in parallel with the sequence program for performing the karaoke performance, and data exchange with the sequence program is also performed. In the following description, it is assumed that data of the song selection number of the karaoke song is received from the control unit 12.

  First, the voice control unit 24 determines whether or not the NG word data transmitted from the control unit 12 has been received (S210). When NG word data is received (S210: YES), the received NG word data is stored in a memory (not shown) of the voice control unit 24, and the process of S211 is executed.

  In the process of S211, the voice control unit 24 determines whether or not the karaoke performance start signal transmitted from the control unit 12 has been received. When the karaoke performance start signal is received (S211: YES), the music data corresponding to the music selection number of the karaoke music is reproduced and the karaoke performance is started (S212).

  Next, the voice control unit 24 determines whether or not the voice recognition start signal transmitted from the control unit 12 has been received (S214). When the voice recognition start signal is received (S214: YES), voice recognition of the lyrics of the karaoke song is started (S216). Specifically, the voice control unit 24 controls the voice recognition unit 10 to recognize the karaoke song lyrics. Then, the lyrics of the karaoke song that has been voice-recognized are stored in the karaoke song lyrics recording area 60 provided in the ROM built in the voice control unit 24.

  And the audio | voice control part 24 judges whether the area song scoring start signal transmitted from the control part 12 was received (S218). When the section singing start signal is received (S218: YES), the reference counter (entire) is initialized (S220).

  Next, it is determined whether or not it is a period in which the prelude period or the interlude period is not sung (S222, see FIG. 10). This determination of S222 is made by determining the prelude period or interlude by section division data indicating the predecessor period, the singing period, and the start point of the interlude period of the karaoke song stored in the control data track delivered from the sequence program that performs the karaoke performance. It is determined whether or not it is a period, that is, a period during which singing is not performed. And when it is not a prelude period or an interlude period, ie, in the case of a singing period (S222: NO), the singing scoring comparison process mentioned later is performed (S224). On the other hand, in the case of a prelude period or an interlude period (S222: YES), a singing scoring adjustment process described later is executed (S226) and a voice recognition adjusting process described later is executed (S228). When the singing scoring comparison process is executed (S224), the singing scoring adjustment process is executed (S226), or the voice recognition adjustment process is executed (S228), the section singing is finished. Whether or not (S230). And when section song scoring is not completed (S230: NO), it returns to S222 and performs the processing mentioned above.

  On the other hand, when the section singing is finished (S230: YES), the pitch difference stored in the difference data storage area 66 of the ROM (not shown) built in the voice control unit 24 by the singing scoring comparison process described later. The entire pitch difference data is extracted from the data (S232), and the total pitch difference data is divided by the reference counter value to average the pitch difference data (S234). Then, using the singing score information (see FIG. 15) corresponding to the pitch difference data stored in the scoring information memory area 54 of the HDD 16, singing is scored according to the pitch difference data (S236). When the process of S236 is completed, the process of S238 (see FIG. 11) is executed.

  In the process of S238, it is determined whether or not the result of the singing score is 80 points or more. The score for judging the result of singing is not limited to “80 points”. It is better to set the score that can prevent the speech recognition rate from being lowered. And when the result of a singing score is not 80 points or more (S238: NO), the process of S254 is performed. On the other hand, when the result of the singing score is 80 points or more (S238: YES), the process of S240 is executed.

  In the process of S240, the speech-recognized lyrics data is compared with the NG word data. Specifically, the lyrics data of the karaoke song that has been voice-recognized and stored in the lyrics recording area 60 of the karaoke song provided in the ROM built in the voice control unit 24 is read. Further, NG word data stored in a memory (not shown) included in the voice control unit 24 is read out. Then, the read lyrics data of the karaoke song is compared with the NG word data. Then, the comparison result, for example, the quantity of NG word data in the lyrics data of the karaoke song is transmitted to the control unit 12 (S242). When the process of S242 is completed, the process of S244 is executed.

  Then, the voice control unit 24 determines whether or not the karaoke performance stop signal transmitted from the control unit 12 has been received (S244). If the karaoke performance does not end (S244: NO), the process returns to S218 (see FIG. 9) to execute the above-described processing. On the other hand, when the karaoke performance stop signal is received (S244: YES), the karaoke performance is stopped (S246). When the process of S246 is completed, the process of S248 is executed.

  Next, the voice control unit 24 determines whether or not the voice recognition end signal transmitted from the control unit 12 has been received (S248). When the voice recognition end signal is received (S248: YES), the voice recognition of the lyrics of the karaoke song is finished (S250). Then, when the process of S250 is completed, the process of S252 is executed.

In the process of S252, NG word singing data is transmitted to the control unit 12.
Then, when the processing of S252 is completed, the present “sound control unit 24 NG word game processing” ends.

  In the process of S254, it is determined whether or not the karaoke performance has ended. If the karaoke performance does not end (S254: NO), the process returns to S218 (see FIG. 9), and the above-described processing is executed. On the other hand, when the karaoke performance is completed (S254: YES), a karaoke performance end signal is transmitted to the control unit 12 (S256).

  Next, the voice control unit 24 determines whether or not the voice recognition end signal transmitted from the control unit 12 has been received (S258). When the voice recognition end signal is received (S258: YES), the voice recognition of the lyrics of the karaoke song is finished (S260).

Then, when the process of S260 is completed, the present “sound control part 24 NG word game process” ends.
In the above scoring, the singing skill is judged by comparing the pitch data. That is, the volume level is not reflected in the score. This is because the sound volume varies widely depending on the vocabulary, gender, age, etc., so that it may take a point far from the skill of actual singing. Also, by not using the volume for scoring, the algorithm can be simplified and the accuracy of singing can be determined in a short time.

(1) Explanation of Singing Score Comparison Processing FIG. 12 is a flowchart showing data fetching processing.
First, FIG. 12A shows a procedure of data capture processing executed by the microphone 25 and the voice control unit 24.

  An analog audio signal output from the microphone 25 to which the singing voice is input is converted into a digital audio signal, and a scoring signal is obtained by subtracting a reproduction signal of music data reproduced by karaoke performance from the audio signal. Generate (S2240). Then, the voice control unit 24 counts the frequency (S2242) in units of 50 ms frames using the scoring signal. The calculated frequency count value is read every 50 ms. This point will be described later with reference to a flowchart showing the singing scoring comparison process of FIG.

  FIG. 12B is a flowchart showing a procedure of guide melody data fetching processing executed by the voice control unit 24. This process is executed when the event data of the guide melody track is delivered from the sequence program that executes the karaoke performance. First, the guide melody data passed from the sequence program is taken into the guide melody buffer 62 of the ROM (not shown) built in the voice control unit 24 (S2244). Pitch data (that is, reference) is extracted from the guide melody data (S2246). Then, the reference data register 64 of the ROM (not shown) built in the voice control unit 24 is updated with the pitch data extracted in this way (S2248). Therefore, the reference data register 64 is updated each time new guide melody data is input.

  Next, FIG. 13 is a flowchart showing the details of the singing scoring comparison process (S224 in FIG. 10) of the “NG word game process of the voice control unit 24” executed by the voice control unit 24. In this process, the frequency count value of the scoring signal is taken and converted into pitch data and frequency data of the scoring signal, and compared with the pitch data of the reference data obtained by the reference data input operation of FIG. Thus, the difference data is obtained. The singing scoring comparison operation is executed every 50 ms, which is one frame time of the singing voice signal.

  First, it is determined whether or not the reference is updated (S2250). If the reference is not updated (S2250: NO), the process returns. When the reference is updated (S2250: YES), the reference counter (entire) is incremented (S2252). Then, the frequency count value described above is read (S2254), and pitch data is generated based on the frequency count value (S2256). Next, the scoring signal and the reference pitch data are compared to calculate the difference (S2258), and the calculated difference is stored as pitch difference data in the storage area corresponding to the current section of the difference data storage area 66. (S2259).

(2) Explanation of Singing Score Adjustment Processing Next, details of the above-described singing score adjustment processing (S226 in FIG. 10) will be described based on the flowchart of FIG.

  First, an analog audio signal output from the microphone 25 to which the singing voice is input is converted into a digital audio signal, and the reproduction signal of the music data reproduced by the karaoke performance is subtracted from the audio signal for scoring. A signal is generated (S2260). Then, the voice control unit 24 detects the signal level of the scoring signal (S2262). Next, it is determined whether or not the signal level of the scoring signal is equal to or higher than a specified value (S2264). When the signal level of the scoring signal is equal to or higher than the specified value (S2264: YES), the reproduction signal of the music data reproduced by the karaoke performance is amplified (S2266). Specifically, as illustrated in FIG. 1, the singing variable gain amplifier M32 is instructed to adjust the gain so that the signal level of the scoring signal is below a specified value.

  On the other hand, if the signal level of the scoring signal is not equal to or higher than the specified value (S2264: NO), the process returns. The singing scoring adjustment process is executed every 50 ms, which is one frame time of the singing voice signal. Therefore, control is performed so that the signal level of the scoring signal becomes equal to or less than the specified value every 50 ms.

(3) Explanation of Adjustment Processing for Speech Recognition Next, details of the above-described speech recognition adjustment processing (S228 in FIG. 10) will be described based on the flowchart in FIG.

  First, an analog audio signal output from the microphone 25 to which the singing voice is input is converted into a digital audio signal, and the audio signal is subtracted from the audio signal to reproduce the music data reproduced by the karaoke performance. A signal for use is generated (S2280). Then, the voice control unit 24 detects the signal level of the voice recognition signal (S2282). Next, it is determined whether the signal level of the voice recognition signal is equal to or higher than a specified value (S2284). If the signal level of the voice recognition signal is equal to or higher than the specified value (S2284: YES), the reproduction signal of the music data reproduced by the karaoke performance is amplified (S2286). Specifically, as illustrated in FIG. 1, the gain is instructed to the lyrics variable gain amplifier M72, and the signal level of the speech recognition signal is adjusted to be equal to or lower than a specified value.

  On the other hand, if the signal level of the voice recognition signal is not equal to or higher than the specified value (S2284: NO), the process returns. The adjustment process of the voice recognition is executed every 50 ms that is one frame time of the singing voice signal. Therefore, control is performed so that the signal level of the speech recognition signal becomes equal to or less than the specified value every 50 ms.

[Description of effects]
(1) In the case of recognizing karaoke singing lyrics in a conventional karaoke device, the karaoke performance sound is emitted from the karaoke device. Therefore, there has been a problem that the recognition rate of speech recognition is low because the karaoke performance sound itself other than the voice for singing karaoke becomes noise.

  On the other hand, according to the karaoke apparatus 1 of the present embodiment, the lyrics signal generation unit M70 compares the sound signal of the karaoke performance sound with the singing sound signal, and subtracts the sound signal of the karaoke performance sound from the singing sound signal. Since the voice recognition signal is generated, the sound signal corresponding to the karaoke performance sound emitted from the speaker M20 is reduced. Then, since the voice recognition unit M76 recognizes the lyrics of the karaoke song based on the voice recognition signal generated by the lyrics signal generation unit M70, an appropriate voice recognition result can be obtained.

  (2) Also, according to the karaoke apparatus 1 of the present embodiment, NG word data is selected from the music data stored in the HDD 16, and the video control unit 38 is controlled to display the selected NG word data. When the lyric data of the karaoke song recognized by the voice recognition unit 10 is compared with the selected NG word data and determined to have the same lyric data, the voice control unit 24 is displayed. Control to stop playback of karaoke performance.

Therefore, according to such a karaoke apparatus 1, it is possible to enjoy an NG word game that competes whether or not an NG word is sung during karaoke singing.
(3) Also, according to the karaoke apparatus 1 of the present embodiment, the lyrics data of the karaoke song recognized by the voice recognition unit 10 and the selected NG word data when the section singing score is a predetermined value, for example, 80 points or more. And compare.

  Therefore, according to such a karaoke apparatus 1, as a condition for playing an NG word game in which NG words are sung during karaoke singing, the singing scoring result is limited to a predetermined value or more, for example, 80 points or more. Thus, it is possible to prevent a decrease in the recognition rate of voice recognition in a game with higher competitiveness.

[Other embodiments]
As mentioned above, although embodiment of this invention was described, this invention is not limited to the said embodiment, It is possible to implement in the following various aspects.

  (1) In the above embodiment, for each section singing score, when the section singing score is a predetermined value, for example, 80 points or more, the lyrics data of the karaoke song recognized by the voice recognition unit 10 and the selected NG word data are In comparison, when the NG word is sung, the game stops the karaoke performance, but the present invention is not limited to this. For each section singing score, when the section singing score is a predetermined value, for example, 80 points or more, the lyrics data of the karaoke song recognized by the voice recognition unit 10 until the end of the karaoke performance and the word data memory area 52 provided in the HDD 16 The stored word data may be compared and the lyrics may be scored based on the quantity of the different lyric data and the word data.

  Below, “the lyrics scoring process of the control unit 12” executed by the control unit 12 of the karaoke apparatus 1 and “the lyrics scoring process of the voice control unit 24” executed by the voice control unit 24 of the karaoke apparatus 1 will be described in order.

[Description of Lyric Scoring Process of Control Unit 12]
Below, the procedure of "the lyrics scoring process of the control part 12" which the control part 12 of the karaoke apparatus 1 performs is demonstrated based on the flowchart of FIG. 17, FIG.

  In the following description, it is assumed that a karaoke song is selected by a user (karaoke singer). Specifically, the music selection number data of the karaoke song received by the operation unit 18 is transmitted to the control unit 12 by the operation processing unit 22, but the control unit 12 receives the music selection number data and receives the music selection number. It is assumed that the number data is being transmitted to the voice control unit 24.

  The lyrics scoring instruction data received by the operation unit 18 is transmitted to the control unit 12 by the operation processing unit 22, but the control unit 12 determines whether or not the lyrics scoring instruction data has been received (S310). . When the lyrics scoring instruction data transmitted from the operation processing unit 22 is received (S310: YES), music data corresponding to the music selection number of the selected karaoke music is provided in the HDD 16 (a). Is read from the music data memory area 50 illustrated in FIG. Then, the word data used in the lyrics illustrated in FIG. 6A is read from the read music data, and the read word data is stored in the word data memory area 52 (S314). When the process of S314 is completed, the process of S320 is executed.

  In the process of S320, a karaoke performance start signal is transmitted to the voice control unit 24. Then, a voice recognition start signal is transmitted to the voice control unit 24 (S322). Furthermore, a section singing start signal is transmitted to the voice control unit 24 (S324, see FIG. 18). When the process of S324 is completed, the process of S326 is executed.

  The speech-recognized lyrics data is transmitted from the speech control unit 24 to the control unit 12 (this transmission process will be described later), but the control unit 12 has received the speech-recognized lyrics data. Is determined (S326). And when the lyrics data by which the speech recognition was carried out from the audio | voice control part 24 was received (S326: YES), the process of S328 is performed. On the other hand, when the lyrics data recognized by the voice from the voice control unit 24 is not received (S326: NO), the process of S338 is executed.

  In the process of S328, the speech-recognized lyrics data is compared with the word data. Specifically, the word data stored in the word data memory area 52 is read out, and the quantity of different lyric data is extracted from the read word data and the speech-recognized lyrics data. Then, the lyrics are scored based on the comparison result (S332). Specifically, the lyrics of the karaoke song are scored based on the quantity of the different lyric data extracted in the process of S328 and the quantity of the word data. More specifically, for example, if there are 5 different lyric data quantities and 100 word data quantities, (100-5 pieces) / 100 pieces = 0.95, out of 100 points. You can score 95 points. Further, comprehensive lyrics are scored (S332). Specifically, the lyric score for each section singing score is weighted and averaged to score the overall lyrics. More specifically, for example, there are five different lyric data quantities in the first section singing score, and there are 100 word data quantities, and the lyric data quantity in the second section singing score is three. If there are 100 words and the quantity of word data is 100, the total number of different lyrics data is 8 and the quantity of word data is 200. Therefore, when the first time and the second time are weighted averaged, (200-8 pieces) / 200 pieces = 0.96, and 96 points out of 100 points can be scored with comprehensive lyrics. When the process of S332 ends, the process of S338 is executed.

In the process of S338, the karaoke performance end signal is transmitted from the voice control unit 24 to the control unit 12 (this transmission process will be described later), but the control unit 12 has received the karaoke performance end signal. Determine whether. And when not receiving a karaoke performance end signal (S338: NO), it returns to S324 and performs the process mentioned above. On the other hand, when a karaoke performance end signal is received (S338: YES), a voice recognition end signal is transmitted to the voice control unit 24 (S340). When the process of S340 ends, the process of S342 is executed.

  In the process of S342, the comprehensive lyrics scoring result is displayed on the display unit 36. Specifically, the video control unit 38 is controlled so that, for example, “total lyrics scoring: 96 points” is displayed on the display unit 36.

Then, when the processing of S342 is completed, the present “lyric scoring processing of the control unit 12” ends.
[Explanation of the lyrics scoring process of the voice control unit 24]
Next, the procedure of “the lyrics scoring process of the voice control unit 24” executed by the voice control unit 24 of the karaoke apparatus 1 will be described based on the flowcharts of FIGS. The operation program related to the lyric scoring process is executed in parallel with the sequence program for performing the karaoke performance, and data exchange with the sequence program is also performed. In the following description, it is assumed that data of the song selection number of the karaoke song is received from the control unit 12.

  First, the voice control unit 24 determines whether or not the karaoke performance start signal transmitted from the control unit 12 has been received (S410). And if a karaoke performance start signal is received (S410: YES), the music data corresponding to the music selection number of a karaoke music will be reproduced | regenerated and a karaoke performance will be started (S412).

  Next, the voice control unit 24 determines whether or not the voice recognition start signal transmitted from the control unit 12 has been received (S414). When the voice recognition start signal is received (S414: YES), voice recognition of the lyrics of the karaoke song is started (S416). Specifically, the voice control unit 24 controls the voice recognition unit 10 to recognize the karaoke song lyrics. Then, the lyrics data of the karaoke song that has been voice-recognized is stored in the karaoke song lyrics recording area 60 provided in the ROM built in the voice control unit 24.

  And the audio | voice control part 24 judges whether the area song scoring start signal transmitted from the control part 12 was received (S418). When the section singing start signal is received (S418: YES), the reference counter (entire) is initialized (S420).

  Next, it is determined whether it is a period during which the prelude period or the interlude period is not sung (S422, see FIG. 20). This determination of S422 is made by determining the prelude period or the interlude based on the segment division data indicating the predecessor period, the singing period, and the start point of the interlude period of the karaoke song stored in the control data track delivered from the sequence program that executes the karaoke performance. It is determined whether or not it is a period, that is, a period during which singing is not performed. And when it is not a prelude period or an interlude period, ie, in the case of a singing period (S422: NO), a singing scoring comparison process is performed (S424). Specifically, the process described in “(1) Description of singing scoring comparison process” in the above embodiment is executed. On the other hand, in the case of the prelude period or the interlude period (S422: YES), the singing scoring adjustment process is executed (S426) and the voice recognition adjustment process is executed (S428). Specifically, the process described in “(2) Explanation of adjustment process of singing scoring” in the above embodiment is executed, and the process described in “(3) Explanation of adjustment process of voice recognition” in the above embodiment. Execute. When the singing scoring comparison process is executed (S424), the singing scoring adjustment process is executed (S426), or the voice recognition adjustment process is executed (S428), the section singing is finished. It is determined whether or not (S430). And when section song scoring is not complete | finished (S430: NO), it returns to S422 and performs the process mentioned above.

  On the other hand, when the section singing is completed (S430: YES), the ROM included in the voice control unit 24 by the singing comparison processing described in “(1) Description of singing comparison processing” in the above embodiment. The entire pitch difference data is extracted from the pitch difference data stored in the difference data storage area 66 (not shown) (S432), and the total pitch difference data is divided by the reference counter value to obtain the pitch difference data. Are averaged (S434). Then, using the singing score information (see FIG. 15) corresponding to the pitch difference data stored in the scoring information memory area 54 of the HDD 16, singing is scored according to the pitch difference data (S436). When the process of S436 is completed, the process of S438 (see FIG. 21) is executed.

  In the process of S438, it is determined whether or not the result of the singing score is 80 points or more. The score for judging the result of singing is not limited to “80 points”. It is better to set the score that can prevent the speech recognition rate from being lowered. And when the result of a singing score is not 80 points or more (S438: NO), the process of S444 is performed. On the other hand, when the result of the singing score is 80 points or more (S438: YES), the process of S442 is executed.

In the processing of S442, the speech-recognized lyrics data is transmitted to the control unit 12. Specifically, the lyrics data of the karaoke song stored in the karaoke song lyrics recording area 60 provided in the ROM built in the voice control unit 24 is read, and the read lyrics data of the karaoke song is sent to the control unit 12. Send.
When the process of S442 ends, the process of S444 is executed.

  In the process of S444, it is determined whether or not the karaoke performance has ended. If the karaoke performance does not end (S444: NO), the process returns to S418 (see FIG. 19) and the above-described processing is executed. On the other hand, when the karaoke performance is completed (S444: YES), a karaoke performance end signal is transmitted to the control unit 12 (S446).

  Next, the voice control unit 24 determines whether or not the voice recognition end signal transmitted from the control unit 12 has been received (S448). When the voice recognition end signal is received (S448: YES), the voice recognition of the lyrics of the karaoke song is ended (S450).

Then, when the processing of S450 is completed, the present “lyric scoring processing of the voice control unit 24” is completed.
By executing the “lyric scoring process of the control unit 12” executed by the control unit 12 of the karaoke apparatus 1 and the “lyric scoring process of the voice control unit 24” executed by the voice control unit 24 of the karaoke apparatus 1 as described above. The lyrics data of the karaoke song recognized by the voice recognition unit 10 until the end of the karaoke performance is compared with the word data stored in the word data memory area 52 provided in the HDD 16, and the quantity of the different lyrics data The lyrics of the karaoke song can be scored based on the word data.

  Moreover, according to the karaoke apparatus 1 of this embodiment, when the section singing score is a predetermined value, for example, 80 points or more, the lyrics data of the karaoke song recognized by the voice recognition unit 10 is compared with the word data.

  Therefore, according to such a karaoke apparatus 1, as a condition for scoring the lyrics of the karaoke song, the singing scoring result is limited to a predetermined value or more, for example, 80 points or more, so that the recognition rate of speech recognition in lyric scoring is reduced. Can be prevented.

  (2) For each section singing score, when the section singing score is a predetermined value, for example, 80 points or more, the lyrics data of the karaoke song recognized by the voice recognition unit 10 until the end of the karaoke performance and the words provided in the HDD 16 If the word data stored in the data memory area 52 is compared and it is determined that there are a predetermined number or more of different lyric data, the game may be stopped.

  Hereinafter, “the lyrics error game process of the control unit 12” executed by the control unit 12 of the karaoke apparatus 1 and “the lyrics error game process of the voice control unit 24” executed by the voice control unit 24 of the karaoke apparatus 1 will be described in order. .

[Explanation of Lyric Error Game Process of Control Unit 12]
Below, the procedure of "the lyrics error game process of the control part 12" which the control part 12 of the karaoke apparatus 1 performs is demonstrated based on the flowchart of FIG. 22, FIG.

  In the following description, it is assumed that a karaoke song is selected by a user (karaoke singer). Specifically, the music selection number data of the karaoke song received by the operation unit 18 is transmitted to the control unit 12 by the operation processing unit 22, but the control unit 12 receives the music selection number data and receives the music selection number. It is assumed that the number data is being transmitted to the voice control unit 24.

  The game start instruction data received by the operation unit 18 is transmitted to the control unit 12 by the operation processing unit 22, but the control unit 12 determines whether or not the game start instruction data has been received (S510). . When the game start instruction data transmitted from the operation processing unit 22 is received (S510: YES), the mathematical expression “n = 0” is stored in the lyrics error number counter (not shown) of the control unit 12 (S511). ). When the process of S511 is completed, the process of S512 is executed.

  In the process of S512, the music data corresponding to the music selection number of the selected karaoke music is read from the music data memory area 50 illustrated in FIG. Then, the word data used in the lyrics illustrated in FIG. 6A is read from the read music data, and the read word data is stored in the word data memory area 52 (S514). When the process of S514 is completed, the process of S520 is executed.

  In the process of S520, a karaoke performance start signal is transmitted to the voice control unit 24. Then, a voice recognition start signal is transmitted to the voice control unit 24 (S522). Furthermore, a section singing start signal is transmitted to the voice control unit 24 (S524, see FIG. 23). When the process of S524 is completed, the process of S526 is executed.

  The speech-recognized lyrics data is transmitted from the speech control unit 24 to the control unit 12 (this transmission process will be described later), but the control unit 12 has received the speech-recognized lyrics data. Is determined (S526). And when the lyrics data by which the speech recognition was carried out from the audio | voice control part 24 was received (S526: YES), the process of S528 is performed. On the other hand, when the lyrics data recognized by the voice from the voice control unit 24 is not received (S526: NO), the process of S538 is executed.

  In the process of S528, it is determined whether or not the lyrics error number counter exceeds a predetermined number. Here, the predetermined number is set according to the level of each karaoke singer, for example, “1” for a skilled karaoke singer, “3” for an intermediate singer, or “5” for a beginner. It is a numerical value. If the lyrics error number counter does not exceed the predetermined number (S528: NO), the process of S537 is executed. On the other hand, when the lyrics error number counter exceeds the predetermined number (S528: YES), a karaoke performance stop signal is transmitted to the voice control unit 24 (S530). When the process of S530 is completed, the process of S532 is executed.

  In the process of S532, a voice recognition end signal is transmitted to the voice control unit 24. Then, the result of the number of lyrics errors is displayed on the display unit 36 (S536). Specifically, for example, “Lyrics error count: 5” corresponding to the result of the number of lyrics errors is displayed on the display unit 36, and “Still! Sorry” is displayed as shown in FIG. The video control unit 38 is controlled to display on the unit 36.

Then, when the process of S536 is completed, the present “scoring process of the control unit 12” is completed.
In the process of S537, the numerical value “1” is added to the mathematical expression “n = 0” stored in the lyrics error counter (not shown) of the control unit 12 to obtain the mathematical expression “n = 1”. The mathematical expression “n = 1” is stored in a memory (not shown) (S537). When the process of S537 is completed, the process of S538 is executed.

In the process of S538, it is determined whether or not a karaoke performance end signal has been received from the voice control unit 24. If the karaoke performance end signal is not received from the voice control unit 24 (S538: NO), the process returns to S524 and the above-described process is executed. On the other hand, when a karaoke performance end signal is received from the voice control unit 24 (S538: YES)
A voice recognition end signal is transmitted to the voice control unit 24 (S540). Then, the result of the number of lyrics errors is displayed on the display unit 36 (S542). Specifically, for example, “Lyrics error number: 0” corresponding to the result of the number of lyrics errors is displayed on the display unit 36, and “Year! Congratulations” is displayed as illustrated in FIG. The video control unit 38 is controlled to display on the display unit 36.

Then, when the process of S542 is completed, the present “scoring process of the control unit 12” ends.
[Description of Lyric Error Game Processing of Voice Control Unit 24]
Next, the procedure of “the lyrics error game process of the voice control unit 24” executed by the voice control unit 24 of the karaoke apparatus 1 will be described based on the flowcharts of FIGS. The operation program relating to the lyric error game processing is executed in parallel with the sequence program for executing the karaoke performance, and data exchange with the sequence program is also performed. In the following description, it is assumed that data of the song selection number of the karaoke song is received from the control unit 12.

  First, the voice control unit 24 determines whether or not the karaoke performance start signal transmitted from the control unit 12 has been received (S610). And if a karaoke performance start signal is received (S610: YES), the music data corresponding to the music selection number of a karaoke music will be reproduced | regenerated and a karaoke performance will be started (S612).

  Next, the voice control unit 24 determines whether or not the voice recognition start signal transmitted from the control unit 12 has been received (S614). When the voice recognition start signal is received (S614: YES), voice recognition of the lyrics of the karaoke song is started (S616). Specifically, the voice control unit 24 controls the voice recognition unit 10 to recognize the karaoke song lyrics. Then, the lyrics of the karaoke song that has been voice-recognized are stored in the karaoke song lyrics recording area 60 provided in the ROM built in the voice control unit 24.

  And the audio | voice control part 24 judges whether the area song scoring start signal transmitted from the control part 12 was received (S618). When the section singing start signal is received (S618: YES), the reference counter (entire) is initialized (S620).

  Next, it is determined whether or not the period of the prelude period or the interlude period is not sung (see S622, FIG. 25). The determination in S622 is made by determining the prelude period or the interlude based on the segment division data indicating the predecessor period, the singing period, and the start point of the interlude period of the karaoke song stored in the control data track delivered from the sequence program that executes the karaoke performance. It is determined whether or not it is a period, that is, a period during which singing is not performed. And when it is not a prelude period or an interlude period, ie, in the case of a singing period (S622: NO), the comparison process of a singing score is performed (S624). Specifically, the process described in “(1) Description of singing scoring comparison process” in the above embodiment is executed. On the other hand, in the case of the prelude period or the interlude period (S622: YES), the singing scoring adjustment process is executed (S626) and the voice recognition adjustment process is executed (S628). Specifically, the process described in “(2) Explanation of adjustment process of singing scoring” in the above embodiment is executed, and the process described in “(3) Explanation of adjustment process of voice recognition” in the above embodiment. Execute. When the singing scoring comparison process is executed (S624), the singing scoring adjustment process is executed (S626), or the voice recognition adjustment process is executed (S628), the section singing is finished. Whether or not (S630). If the section singing has not been completed (S630: NO), the process returns to S622 and the above-described processing is executed.

  On the other hand, when the section singing is completed (S630: YES), the ROM included in the voice control unit 24 by the singing scoring comparison processing described in “(1) Description of singing scoring comparison processing” in the above embodiment. The entire pitch difference data is extracted from the pitch difference data stored in the difference data storage area 66 (not shown) (S632), and the total pitch difference data is divided by the reference counter value to obtain the pitch difference data. Are averaged (S634). Then, using the singing score information (see FIG. 15) corresponding to the pitch difference data stored in the scoring information memory area 54 of the HDD 16, singing is scored according to the pitch difference data (S636). When the process of S636 is completed, the process of S638 (see FIG. 26) is executed.

  In the process of S638, it is determined whether or not the result of the singing score is 80 points or more. The score for judging the result of singing is not limited to “80 points”. It is better to set the score that can prevent the speech recognition rate from being lowered. And when the result of a singing score is not 80 points or more (S638: NO), the process of S654 is performed. On the other hand, when the result of the singing score is 80 points or more (S638: YES), the process of S642 is executed.

In the process of S642, the lyric data recognized by the voice is transmitted to the control unit 12. When the process of S642 is completed, the process of S644 is executed.
Then, the voice control unit 24 determines whether or not the karaoke performance stop signal transmitted from the control unit 12 has been received (S644). If the karaoke performance does not end (S644: NO), the process returns to S618 (see FIG. 24) and the above-described processing is executed. On the other hand, when the karaoke performance stop signal is received (S644: YES), the karaoke performance is stopped (S646). When the process of S646 ends, the process of S648 is executed.

  Next, the voice control unit 24 determines whether or not the voice recognition end signal transmitted from the control unit 12 has been received (S648). When the voice recognition end signal is received (S648: YES), the voice recognition of the lyrics of the karaoke song is finished (S650).

Then, when the process of S650 is completed, the present “lyric error game process of the voice control unit 24” is completed.
In the process of S654, it is determined whether or not the karaoke performance has ended. If the karaoke performance does not end (S654: NO), the process returns to S618 (see FIG. 24) to execute the above-described processing. On the other hand, when the karaoke performance is completed (S654: YES), a karaoke performance end signal is transmitted to the control unit 12 (S656).

  Next, the voice control unit 24 determines whether or not the voice recognition end signal transmitted from the control unit 12 has been received (S658). When the voice recognition end signal is received (S658: YES), the voice recognition of the lyrics of the karaoke song is finished (S660).

Then, when the processing of S660 is completed, the present “lyric error error game processing of the voice control unit 24” is completed.
The “lyric error error game process of the control unit 12” executed by the control unit 12 of the karaoke apparatus 1 and the “lyric error error game process of the audio control unit 24” executed by the audio control unit 24 of the karaoke apparatus 1 are executed. Thus, even during karaoke singing, if a predetermined number of lyrics are sung by mistake, karaoke performance is stopped, so that the game performance can be improved.

  Moreover, according to the karaoke apparatus 1 of this embodiment, when the section singing score is a predetermined value, for example, 80 points or more, it is determined whether the number of lyrics errors of the karaoke song recognized by the voice recognition unit 10 is a predetermined number or more.

  Therefore, according to such a karaoke apparatus 1, as a condition for playing a game in which karaoke performance is stopped if the lyrics are incorrect, the singing scoring result is limited to a predetermined value or more, for example, 80 points or more. It is possible to prevent a decrease in the recognition rate of voice recognition in a high game.

It is a figure which shows schematic structure centering on the function of a karaoke apparatus. On / off changes of the switch M40, the switch M38 and the switch M78 with respect to the elapsed time after the karaoke performance period, before the karaoke performance period, and the output signal level from the singing variable gain amplifier M32 and the lyrics variable gain amplifier M72 It is explanatory drawing which shows this change. It is explanatory drawing which shows the pitch data of the guide melody data of a music database, and the pitch data of the signal for scoring. 1 is a block diagram illustrating a configuration of a karaoke apparatus 1. FIG. (A) is explanatory drawing which shows the memory area provided in HDD16, (b) is explanatory drawing which shows the structure of MIDI data, (c) is in ROM (not shown) which the audio | voice control part 24 incorporates. It is explanatory drawing which shows the lyrics recording area 60 of the provided karaoke song, the guide melody buffer 62, the reference data register 64, and the difference data storage area 66. (A) is explanatory drawing which shows a music data format, (b) is explanatory drawing which shows the example of the word data used by the lyrics. 7 is a flowchart showing a part of a procedure of “NG word game process of control unit 12” executed by control unit 12. 7 is a flowchart showing a part of a procedure of “NG word game process of control unit 12” executed by control unit 12. 10 is a flowchart showing a part of a procedure of “NG word game process of voice control unit 24” executed by the voice control unit 24; 10 is a flowchart showing a part of a procedure of “NG word game process of voice control unit 24” executed by the voice control unit 24; 10 is a flowchart showing a part of a procedure of “NG word game process of voice control unit 24” executed by the voice control unit 24; (A) is a flowchart which shows the procedure of the data acquisition process performed in [comparison process] of "the singing scoring of the audio | voice control part 24" which the audio | voice control part 24 performs, (b) is the audio | voice control part 24. It is a flowchart which shows the procedure of the taking-in process of the guide melody data performed in [comparison process] of "the singing scoring of the audio | voice control part 24" to perform. It is a flowchart which shows the detail of the comparison process of the singing score of "NG word game process of the voice control part 24" which the voice control part 24 performs. (A) is a flowchart which shows the detail of the adjustment process of the singing score of "NG word game process of the voice control part 24" which the voice control part 24 performs, (b) is the "voice control which the voice control part 24 performs. It is a flowchart which shows the detail of the adjustment process of the singing score of "NG word game process of the part 24." It is explanatory drawing which shows the scoring information according to the data of pitch difference. (A) is explanatory drawing which shows the example which displayed the title on the display part 36, (b) is explanatory drawing which shows the example which displayed the NG word on the display part 36, (c) is NG word lyrics is an explanatory view showing an example of displaying the table radical 113 36 screen, (d) is an explanatory diagram showing an example of displaying on the display unit 36 Congratulations screen, (e) display the sorry screen 36 It is explanatory drawing which shows the example displayed on. It is a flowchart which shows a part of procedure of the "lyric scoring process of the control part 12" which the control part 12 performs. It is a flowchart which shows a part of procedure of the "lyric scoring process of the control part 12" which the control part 12 performs. 10 is a flowchart showing a part of a procedure of “lyric scoring processing of the voice control unit 24” executed by the voice control unit 24; 10 is a flowchart showing a part of a procedure of “lyric scoring processing of the voice control unit 24” executed by the voice control unit 24; 10 is a flowchart showing a part of a procedure of “lyric scoring processing of the voice control unit 24” executed by the voice control unit 24; 12 is a flowchart showing a part of a procedure of “lyric error game process of control unit 12” executed by control unit 12. 12 is a flowchart showing a part of a procedure of “lyric error game process of control unit 12” executed by control unit 12. 10 is a flowchart showing a part of a procedure of “lyric error error game processing of the voice control unit 24” executed by the voice control unit 24; 10 is a flowchart showing a part of a procedure of “lyric error error game processing of the voice control unit 24” executed by the voice control unit 24; 10 is a flowchart showing a part of a procedure of “lyric error error game processing of the voice control unit 24” executed by the voice control unit 24;

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Karaoke apparatus, 2 ... Remote control terminal, 10 ... Voice recognition part, 12 ... Control part, 14 ... Interface part, 16 ... Hard disk (HDD), 18 ... Operation part, 20 ... Infrared communication part, 22 ... Operation processing part, 24 ... Audio control unit, 25 ... Microphone, 28 ... Speaker, 30 ... MIDI sound source, 32 ... Video RAM, 34 ... Video playback unit, 36 ... Display unit, 38 ... Video control unit, 39 ... Bus, 40 ... USB, 50 ... Music data memory area, 52 ... Word data memory area, 54 ... Scoring information memory area, 60 ... Karaoke song lyrics storage area, 62 ... Guide melody buffer, 64 ... Reference data register, 66 ... Difference data storage area, 100 ... network.

Claims (7)

An audio signal input means for inputting an audio signal of karaoke singing;
Song data storage means for storing song data of karaoke songs;
Karaoke performance reproduction means for reproducing the music data stored in the music data storage means as a sound signal, and outputting the reproduced sound signal and the voice signal of the karaoke song input from the voice signal input means to a speaker;
The first signal corresponding to the sound signal output from the karaoke performance reproducing means is compared with the second signal corresponding to the karaoke singing voice signal input from the voice signal input means, and the second signal is First generation means for generating a speech recognition signal obtained by subtracting the first signal;
First gain setting means for setting a gain for the first generation means to subtract the first signal from the second signal based on the speech recognition signal generated by the first generation means;
Voice recognition means for recognizing lyrics of a karaoke song based on the voice recognition signal generated by the first generation means;
The first signal corresponding to the sound signal output from the karaoke performance reproducing means is compared with the second signal corresponding to the karaoke singing voice signal input from the voice signal input means, and the second signal is Second generating means for generating a scoring signal obtained by subtracting the first signal;
Pitch extraction means for extracting pitch data from the scoring signal generated by the second generation means;
Second gain setting means for setting a gain for the second generation means to subtract the first signal from the second signal based on the scoring signal generated by the second generation means;
The pitch difference between the pitch data of the singing melody of the karaoke song stored in the music data storage means and the pitch data of the scoring signal extracted by the pitch extraction means is calculated, and the calculated pitch difference Singing scoring means for scoring karaoke singing songs for each predetermined section based on
With
In the case of a karaoke song performance period, the first gain setting means sets the gain so that the signal level of the speech recognition signal is minimized, and the second gain setting means The gain setting means sets the gain so that the signal level of the scoring signal is minimized.
Karaoke apparatus according to claim. The karaoke apparatus according to claim 1,
The music data storage means stores music data including lyrics data of karaoke music,
Compare the lyrics data of the karaoke song recognized by the voice recognition means and the lyrics data of the karaoke song stored in the music data storage means, extract the quantity of the different lyrics data, and extract the quantity and A karaoke apparatus comprising lyric scoring means for scoring the lyrics of karaoke songs based on the quantity of lyrics data of karaoke songs. The karaoke apparatus according to claim 1,
The music data storage means stores music data including lyrics data of karaoke music,
The lyrics data of the karaoke song recognized by the voice recognition means and the lyrics data of the karaoke song stored in the music data storage means are compared to extract the quantity of the different lyrics data, and the extracted quantity is predetermined. A karaoke apparatus comprising control means for controlling the karaoke performance reproduction means to stop reproduction of the karaoke performance when it is determined that there are more than one. The karaoke apparatus according to claim 1,
It has a display means that can display lyrics data of karaoke songs,
The music data storage means stores music data including lyrics data of karaoke music,
Extracting specific lyrics data from the song data stored in the song data storage means, controlling the display means to display the extracted lyrics data, and recognizing the karaoke song recognized by the voice recognition means If the lyric data is compared with the extracted lyric data and it is determined that there is the same lyric data, the karaoke performance reproducing means is controlled to stop the reproduction of the karaoke performance. A karaoke device that features it. In the karaoke apparatus according to claim 2, further comprising a recognized karaoke singing voice recognition memory means to store the lyrics data of the said speech recognition means,
The lyrics scoring means determines whether or not the singing result of the karaoke singing by the singing scoring means is greater than or equal to a predetermined value for each predetermined section. Read out the lyrics data of the stored karaoke song, compare the read out lyrics data of the karaoke song with the lyrics data of the karaoke song stored in the song data storage means, and extract the quantity of the different lyrics data A karaoke apparatus for scoring the lyrics of a karaoke song based on the extracted quantity and the quantity of lyrics data of the karaoke song. In the karaoke apparatus according to claim 3, further comprising a recognized karaoke singing voice recognition memory means to store the lyrics data of the said speech recognition means,
The control means determines whether or not the singing result of karaoke singing by the singing grading means is greater than or equal to a predetermined value for each predetermined section. The karaoke song lyrics data is read out, the karaoke song lyrics data read out is compared with the karaoke song lyrics data stored in the song data storage means, and the quantity of the different lyric data is extracted, A karaoke apparatus characterized in that when it is determined that the extracted quantity is a predetermined number or more, the karaoke performance reproduction means is controlled to stop reproduction of the karaoke performance. In the karaoke apparatus according to claim 4, further comprising a recognized karaoke singing voice recognition memory means to store the lyrics data of the said speech recognition means,
The control means extracts specific lyric data from music data stored in the music data storage means, controls the display means to display the extracted lyric data, and displays the extracted lyric data for each predetermined section. It is determined whether or not the singing score result of the karaoke singing by the singing scoring means is greater than or equal to a predetermined value. If the lyric data of the karaoke song read is compared with the extracted lyric data and it is determined that there is the same lyric data, the karaoke performance reproducing means is controlled to stop the reproduction of the karaoke performance A karaoke apparatus characterized by that.