JP6236807B2 - Singing voice evaluation device and singing voice evaluation system - Google Patents

Description

  The present invention relates to a singing voice evaluation device and a singing voice evaluation system for evaluating a singing voice.

  Conventionally, some karaoke apparatuses have a function of scoring a singer's singing. The scoring in the karaoke device is generally performed by converting the similarity between the singing voice and the guide melody into a score, as compared to a guide melody as a model.

  However, such a machine scoring increases the score when it resembles a guide melody, and sensory evaluation of the singing itself has not been made. For example, when a song is sung with an arrangement, the score may be low because it is not similar to a guide melody even if the song is perceived as good by humans.

  Therefore, for example, in Patent Document 1, a singing evaluation is performed by a plurality of listeners, and the higher score between the average score of the listeners and the machine score is used as a scoring result, so that the human subjective Karaoke devices that reflect singing evaluation have been proposed.

JP-A-8-115091

  However, in the apparatus of Patent Document 1, it is necessary to wait for singing evaluation by a plurality of listeners, and the result cannot be presented on the spot.

  Therefore, an object of the present invention is to provide a singing voice evaluation device capable of presenting singing evaluation by a human on the spot.

  The singing voice evaluation apparatus according to the present invention includes a storage unit that stores in advance a singing voice and a human evaluation of the singing voice and stores the singing voice data in advance, a singing voice input unit that inputs the singing voice, and the singing voice input. Scoring means for scoring the current singing voice input by the means. And the scoring means compares the current singing voice with the pre-singing voice data, extracts pre-singing voice data similar to the current singing voice, and human evaluation in the extracted pre-singing voice data Is included in the scoring result and output.

  As described above, the singing voice evaluation device of the present invention stores the singing voice associated with the human evaluation in advance, and extracts the singing voice similar to the current singing voice, so that the similar singing voice is obtained. Can be reflected in the scoring results. For example, in the pre-singing voice data that was sung by changing the pitch by adding an arrangement, it is possible to change the pitch in the same arrangement when the human feeling is good and a high human evaluation is obtained. Singing, the associated human evaluation is reflected in the scoring result, so that the human evaluation can be presented on the spot.

  In addition, not only one similar pre-singing voice data but also a plurality are extracted, each human evaluation is weighted according to the degree of similarity of each extracted pre-singing voice data, and reflected in the scoring result Is preferred.

  The scoring results preferably reflect not only human evaluation but also machine scoring results. Moreover, it is preferable that the result of machine scoring is also included in the pre-song voice data, and the result of machine scoring included in the pre-song voice data is reflected in the scoring result.

  As the scoring result, only the result of one song may be output, or may be output for each predetermined section (for example, A melody, B melody, chorus).

  In addition, this invention performs a singing in a singer's terminal (a karaoke apparatus or the information processing apparatus which a user owns), the said singer's terminal downloads prior singing voice data from the memory | storage means of a server, and the said scoring process A mode of performing (scoring means) is also possible.

  According to this invention, singing evaluation by a human can be presented on the spot.

It is a block diagram which shows the structure of a karaoke system. It is a block diagram which shows the structure of a karaoke apparatus. It is a figure which shows the structure of music data. It is a figure explaining the concept of machine scoring. It is a block diagram which shows the structure of the server 1 and the evaluator terminal 4. FIG. 6 (A) is a diagram showing pre-song audio data, FIG. 6 (B) is a diagram showing list data of pre-song audio data, and FIG. 6 (C) is a human evaluation utilization scoring system. It is a figure which shows a concept. It is a figure which shows the concept of similarity calculation. It is a figure which shows the example 1 of application of human evaluation utilization scoring. It is a figure which shows the example 2 of application of human evaluation utilization scoring. It is a flowchart which shows the operation | movement of a human evaluation utilization scoring process.

  FIG. 1 is a diagram showing a configuration of a karaoke system provided with a singing voice evaluation apparatus of the present invention. The karaoke system includes a server 1 connected via a network 2 such as the Internet, a plurality of karaoke stores 3, and a plurality of evaluator terminals 4.

  Each karaoke store 3 is provided with a relay device 5 such as a router connected to the network 2 and a plurality of karaoke devices 7 connected to the network 2 via the relay device 5. The repeater 5 is installed in a management room of the karaoke store 3 or the like. A plurality of karaoke apparatuses 7 are installed in each private room (karaoke box). Each karaoke device 7 is provided with a remote controller 9.

  The evaluator terminal 4 is an information processing terminal such as a PC or a smartphone. The user of the evaluator terminal 4 evaluates the pre-song voice data (results of singing by the singer at each karaoke device 7) stored in the server 1. In this embodiment, if each karaoke apparatus 7 sings, the singing voice of the said singer will be transmitted to the server 1, and it will become prior singing voice data waiting for human evaluation. The user of the evaluator terminal 4 listens to the singing voice of the pre-singing voice data and gives a score. This personal evaluation is registered in the pre-singing voice data and accumulated in the server 1. And when the singing voice of the singer who sang the same song later is transmitted to the server 1, the server 1 compares the received current singing data with the pre-singing voice data, and extracts similar pre-singing voice data. Then, the human evaluation in the extracted prior singing voice data is reflected in the scoring result in the current singing voice.

  FIG. 2 is a block diagram showing a configuration of the karaoke apparatus 7. The karaoke apparatus 7 includes a CPU 11 that controls the operation of the entire apparatus, and various components connected to the CPU 11. The CPU 11 includes a RAM 12, an HDD 13, a network interface (I / F) 14, an LCD (touch panel) 15, an A / D converter 17, a sound source 18, a mixer (effector) 19, a decoder 22 such as an MPEG, a display processing unit 23, an operation The unit 25 and the transmission / reception unit 26 are connected.

  The HDD 13 stores an operation program for the CPU 11. The RAM 12, which is a work memory, has an area for reading out the operation program for the CPU 11, an area for reading out song data for playing karaoke songs, an area for temporarily storing data such as a reservation list and scoring results, and the like. Is set. The HDD 13 stores music data for playing karaoke music. Further, the HDD 13 also stores video data for displaying a background video on the monitor 24. Video data stores both moving images and still images. Music data and video data are periodically distributed from the server 1 (or other distribution center) and updated.

  The CPU 11 is a control unit that comprehensively controls the karaoke apparatus 7 and functionally incorporates a sequencer to perform karaoke performance. The CPU 11 performs audio signal generation processing, video signal generation processing, machine scoring processing, and human evaluation use scoring processing.

  The touch panel 15 and the operation unit 25 are provided on the front surface of the karaoke apparatus 7. The CPU 11 displays an image corresponding to the operation information on the touch panel 15 based on the operation information input from the touch panel 15 to realize a GUI. The remote controller 9 also realizes the same GUI. The CPU 11 performs various operations based on operation information input from the remote controller 9 via the touch panel 15, the operation unit 25, or the transmission / reception unit 26. For example, when the user gives an instruction to start the human evaluation use scoring using the touch panel 15, the operation unit 25, or the remote controller 9, the CPU 11 starts the human evaluation use scoring process. Details of the human evaluation use scoring process will be described later.

  Next, a configuration for performing karaoke performance will be described. As described above, the CPU 11 functionally includes a sequencer. The CPU 11 reads music data corresponding to the music number of the reserved music registered in the reserved list in the RAM 12 from the HDD 13, and performs a karaoke performance with the sequencer.

  For example, as shown in FIG. 3, the music data includes a header in which a music number is written, a musical sound track in which performance MIDI data is written, a guide melody track in which MIDI data for guide melody is written, and lyrics It consists of a lyrics track in which MIDI data is written, a back chorus playback timing, a chorus track in which audio data to be played back is written, and the like. Note that the format of the music data is not limited to this example.

  In the musical sound track, information indicating the type, timing, pitch (key), strength, length, localization (pan), sound effect (effect), etc. of the musical instrument that generates the musical sound is recorded. The sequencer controls the sound source 18 based on the data of the musical tone track and the guide melody track, and generates the musical tone of the karaoke song.

  The sequencer also reproduces the back chorus audio data (compressed audio data such as MP3 attached to the music data) at the timing designated by the chorus track. Further, the sequencer synthesizes the character pattern of the lyrics in synchronism with the progress of the song based on the lyrics track, converts the character pattern into a video signal, and inputs it to the display processing unit 23.

  The sound source 18 forms a musical sound signal (digital audio signal) according to data (note event data) input from the CPU 11 by processing of the sequencer. The formed tone signal is input to the mixer 19.

  The mixer 19 gives an acoustic effect such as an echo to the musical sound signal generated by the sound source 18, the chorus sound, and the singing voice of the singer input from the microphone 16 via the A / D converter 17. Mix the signal.

  Each mixed digital audio signal is input to a sound system (SS) 20. The sound system 20 incorporates a D / A converter and a power amplifier, converts an input digital signal into an analog signal, amplifies it, and emits sound from the speaker 21. The effect that the mixer 19 gives to each audio signal and the balance of mixing are controlled by the CPU 11.

  The CPU 11 reads the video data stored in the HDD 13 and reproduces the background video and the like in synchronism with the generation of musical sounds by the sequencer and the generation of the lyrics telop. The video data of the moving image is encoded in the MPEG format.

  The CPU 11 inputs the read video data of the background video to the decoder 22. The decoder 22 converts the input data such as MPEG into a video signal and inputs it to the display processing unit 23. In addition to the video signal of the background video, the display processing unit 23 receives a video such as the character pattern of the lyrics telop. The display processing unit 23 combines the video of the lyrics telop with the OSD on the video signal of the background video and outputs it to the monitor 24. The monitor 24 displays the video signal input from the display processing unit 23.

  Karaoke performances are performed as described above. Next, the machine scoring process will be described. The machine scoring process is performed by comparing the singing voice of the singer with the guide melody track that is the reference singing voice. The machine scoring process is performed by comparing the pitch (pitch), timing, volume, etc. for each note event data of the guide melody track.

  That is, the CPU 11 temporarily stores the input singing voice (digital voice signal) in the RAM 12 and extracts the pitch of the singing voice. The CPU 11 compares the extracted pitch value, the timing at which the pitch changes, the level of the singing voice, and the like with the guide melody track, and converts it into a score.

  For example, when the pitch of the singing voice matches the pitch of the guide melody track for a predetermined time or longer (within an allowable range), a high score is given. Also, the timing of pitch change is taken into consideration in the score. Furthermore, points are also awarded based on the presence or absence of techniques such as vibrato, inflection, and sneezing (moving gently from a low pitch).

  For example, as shown in FIG. 4, in the section of note A, the pitch of the singing voice matched the pitch of the guide melody track for a predetermined time or longer (it was within the allowable range), so the score for that section is 70. Points are given. On the other hand, in the section of note B, since the pitch of the singing voice and the pitch of the guide melody track are greatly different, a low score (20 points) is given as the score of the section. In addition, in the section of note C, vibrato skill was detected in addition to 70 points as the basic score, so a total of 90 points are given as the score of the section.

  The score for each note as described above is totaled for each predetermined section (or every elapse of a predetermined time) such as A melody, B melody, and chorus. The total score is temporarily stored in the RAM 12 of the own device. In the human evaluation use scoring process, the total score and the singing voice (digital voice signal) are transmitted to the server 1 via the network I / F 14. Even when the performance of one song is finished, the result of the machine scoring through one song and the singing voice (digital voice signal) are transmitted to the server 1. The machine scoring may be performed by the server 1 instead of each karaoke device 7. In this case, only the singing voice is transmitted to the server 1. Further, in the machine scoring, the process for scoring as described above is not essential. For example, a mode in which information such as a pitch shift amount with respect to the guide melody and timing shift amount is transmitted to the server 1 and the information is stored in the server 1 may be used.

  Next, the human evaluation use scoring process will be described. The human evaluation use scoring process is mainly performed by the server 1. FIG. 5A is a block diagram showing the configuration of the server 1.

  The server 1 is an information processing apparatus that includes a CPU 51, a RAM 52, an HDD 53, and a network I / F 54. The CPU 51 reads out the operation program of the server 1 stored in the HDD 53 to the RAM 52, and performs a scoring result accumulation process and a singing voice comparison process.

  Pre-singing voice data is stored in the HDD 53 of the server 1 (corresponding to the storage means of the present invention). FIG. 6A is a diagram showing the structure of the pre-singing voice data. The singing voice (digital voice signal) and the result of the machine scoring are transmitted to the server 1 from each karaoke device 7 which has instructed start of scoring using human evaluation. The server 1 performs a singing voice comparison process and a scoring result accumulation process using the received singing voice. The singing voice comparison process will be described later, and the scoring result accumulation process will be described first.

  The CPU 51 of the server 1 associates the received singing voice with the result of the machine scoring, assigns a predetermined header (data name, song number, etc.) and stores it in the HDD 53. In addition, other information, such as information on a singer's name, may be received and information, such as a singer's name, may also be stored in association with each other. The score evaluated by the user of the evaluator terminal 4 is stored as the average human score in the pre-singing voice data. However, when the data is first stored in the HDD 53, there is no average human score. In this way, the pre-singing voice data as shown in FIG.

  Next, FIG. 5B is a block diagram showing a configuration of the evaluator terminal 4. The evaluator terminal 4 is an information processing apparatus including a CPU 71, a RAM 72, a ROM 73, a network I / F 74, a display processing unit 75, a monitor 76, an operation unit 77, a sound system (SS) 78, and a speaker 79.

  The CPU 71 reads out the operation program stored in the ROM 73 to the RAM 72 and performs a singing voice reproduction process and a human evaluation process.

  The user of the evaluator terminal 4 gives an instruction to start human evaluation using the operation unit 77. When the CPU 71 receives an instruction to start the personal evaluation via the operation unit 77, the CPU 71 transmits the start instruction to the server 1 via the network I / F 74. When the server 1 receives the start instruction, the server 1 transmits list data of the pre-singing voice data to the evaluator terminal 4.

  As shown in FIG. 6B, the list data includes, for example, the data name of each pre-song audio data, the song name corresponding to the song number, the singer name, the average human scoring result, the machine scoring result, and the like. . The CPU 71 of the evaluator terminal 4 receives the list data from the server 1 and generates video data corresponding to the list data. The CPU 71 outputs the video data to the display processing unit 75 and displays it on the monitor 76. The user selects a singing voice to be reproduced from the list via the operation unit 77.

  When the singing voice is selected, the CPU 71 transmits a reproduction request for the selected singing voice to the server 1. The CPU 51 of the server 1 reads the pre-song audio data corresponding to the reproduction request and transmits it to the evaluator terminal 4.

  The CPU 71 of the evaluator terminal 4 reproduces the singing voice (digital voice signal) of the received prior singing voice data and outputs it to the sound system 78. The sound system 78 converts the input singing voice (digital signal) into an analog signal, amplifies it, and emits the sound from the speaker 79. In this way, the singing voice reproduction process is performed.

  The user listens to the reproduced singing voice and scores through the operation unit 77. The scoring may be in the form of giving a score as a whole song, or in the form of giving a score for each predetermined section such as A melody and B melody. The result of scoring by the user (result of human evaluation) is transmitted to the server 1.

  The CPU 51 of the server 1 averages the received score with the score accumulated in the past, and stores it in the pre-song voice data as an average human score. The average human scoring may be stored as a score for the entire song, or may be stored as a score for each predetermined section such as A melody and B melody. In this way, scoring result accumulation processing is performed. It should be noted that the higher the number of times scored, the more popular the singing voice, and the higher the average human score may be. Moreover, you may perform the weighted average process which provided different weighting for every user who scores. For example, the weighting of scores scored by a user who scores a large number of times is increased.

  Next, the singing voice comparison process will be described. The singing ability comparison process is a process of comparing the singing voice transmitted from the karaoke apparatus 7 with the pre-singing voice data and calculating the similarity. The CPU 51 of the server 1 temporarily stores the singing voice (digital voice signal) received from the karaoke device 7 in the RAM 52 and extracts the pitch of the singing voice. And CPU51 reads the pre-song audio | voice data of the same music as the received singing audio | voice, and compares with the singing audio | voice (digital audio | voice signal) contained in the read pre-singing audio | voice data. In addition, although all the prior song voice data of the same music may be read, for example, only a plurality (for example, ten) may be read from the latest previous song voice data.

  The similarity is calculated by the same method as the machine scoring. That is, the pitch value extracted from the received singing voice, the timing at which the pitch changes, the level of the singing voice, etc., are compared with the pitch, timing, level, etc. of the singing voice included in the pre-singing voice data, This is a process of scoring. However, since note event data such as a guide melody does not exist in the singing voice, a section in which the same pitch continues to some extent among received singing voices (or prior singing voices) is regarded as one note event data for comparison. Do.

  FIG. 7A and FIG. 7B are diagrams showing the concept of similarity calculation. The current singing voice (received singing voice) shown in the figure is the same as the singing voice shown in FIG. FIG. 7 (A) is a diagram comparing with pre-singing voice data when singing faithfully to the pitch of the guide melody.

  The current singing voice in this example is sung by changing the pitch by adding arrangement in the section of note B. Therefore, a lower similarity (20%) is calculated in the section of note B compared to the pre-song voice data when singing faithfully to the pitch of the guide melody.

  On the other hand, FIG. 7 (B) is a figure compared with prior singing voice data in the case of singing by changing the pitch of the section of note B by adding an arrangement. In this example, since both the current singing voice and the pre-singing voice are sung by changing the pitch in the section of note B, the high similarity (70%) is calculated.

  The degree of similarity as described above is totaled for each predetermined section (or every predetermined time) such as A melody, B melody, and rust. Alternatively, the average similarity (total similarity) through one song is calculated.

  The CPU 51 of the server 1 extracts prior singing voice data having a high degree of similarity for each section or for the entire section. And the average human scoring given to the extracted prior singing voice data is reflected in the scoring result of the current singing voice. For example, as shown in FIG. 6 (C), the average human score (80 points) given to the pre-song voice data C with the highest similarity and the pre-song voice data C with the highest similarity are given. The score (70 points) obtained by averaging the machine scoring (60 points) is output as the current singing voice scoring result.

  Alternatively, the average human score (80 points) given to the pre-song voice data C with the highest degree of similarity may be output as a scoring result as it is, or given to the pre-song voice data C with the highest degree of similarity. A score (72.5 points) obtained by averaging the average human score (80 points) and the machine score (for example, 65 points) in the current singing voice may be output as the score result.

  For example, as shown in FIG. 7B, when a high score is given to the average human scoring of the pre-song voice data sung by changing the pitch by adding an arrangement, even if the result of the machine scoring is low, The human evaluation by the arrangement will be reflected, and a scoring result higher than the machine scoring will be obtained. On the other hand, even if the singing was faithful to the guide melody, the singing that the human did not feel good was lower than the mechanical scoring because the human scoring was reflected even though the mechanical scoring result was high A scoring result is obtained.

  As described above, in the karaoke system of this embodiment, when a person who feels good (or poor) and obtained a high (or low) human evaluation is stored in a similar manner (for example, When singing with the pitch changed in the arrangement, the associated high (or low) human evaluation is reflected in the scoring result, so that the human evaluation can be presented on the spot.

  In addition, in the example of FIG. 6C, although the example using the human evaluation of the most similar pre-song voice data was shown, a plurality of pre-song voice data is extracted, and each of the pre-song voice data extracted is extracted. It is preferable to weight each human evaluation according to the degree of similarity and reflect it in the scoring results.

  FIG. 8 is a diagram showing an application example 1 of human evaluation use scoring. The application example 1 shows an example in which a plurality (four) of pre-singing voice data with high similarity are extracted.

  In this example, a weighted addition process is performed on the scoring results according to the degree of similarity between the current singing voice and the pre-singing voice data. That is, the highest weight (contribution rate) is set to the pre-song voice data having the highest similarity, and the scoring result given to the pre-song voice data having the highest similarity is greatly reflected, and the similarity becomes low. Accordingly, the weight (contribution rate) is lowered, and the scoring result is changed according to the similarity.

  Moreover, in this example, the score which averaged the average human scoring and machine scoring in each prior singing voice data is calculated as a composite scoring. Then, the composite scoring and the contribution rate are multiplied to calculate a contribution rate correction point. For example, the preliminary singing voice data C is calculated as a composite score obtained by averaging the average human score (80 points) and the machine score (60 points), but the contribution rate is 40%. 28 points are given as points. Similarly, the preliminary singing voice data B is given 23.55 points as correction points, the preliminary singing voice data A is given 16.2 points as correction points, and the preliminary singing voice data D is 7 points as correction points. .35 points will be awarded. And the score (75.1 points) which added these correction points is output as a scoring result.

  In this example as well, the average human scoring given to each extracted pre-song voice data may be multiplied by the contribution rate and added, so that the mechanical scoring may not be considered. However, by reflecting machine scoring, it is possible to reflect not only subjective evaluation by humans but also objective evaluation such as accuracy of pitch, and more accurate scoring can be performed.

  Next, FIG. 9 is a figure which shows the example 2 of application of human evaluation utilization scoring. The application example 2 shows an example in which the machine scoring result of the current singing voice is also reflected in the application example 1.

  In this example, as the composite scoring, the average human scoring in each pre-song speech data is corrected by the mechanical scoring result of the current singing speech. That is, in this example, a composite score of each preliminary singing voice data is calculated as average human scoring × (current singing machine scoring / pre-singing voice data mechanical scoring) = composite scoring. Then, the contribution rate correction score is calculated by multiplying the calculated composite scoring result by the contribution rate.

  For example, the pre-singing voice data C is obtained by multiplying the average human scoring (80 points) by the current singing machine scoring (65 points) / machine scoring of the pre-singing voice data (60 points), and 80 × (65/60). = 86.67 points are calculated as a composite score. Since the contribution rate is 40%, 34.67 points are given as correction points. Similarly, the pre-singing voice data B is given 18.3 points as correction points, the pre-singing voice data A is given 13.32 points as correction points, and the pre-singing voice data D is 5 points as correction points. .75 points are awarded. And the score (72.04 points) which added these correction points is output as a scoring result.

  In this way, it is possible to perform more accurate scoring by performing fine adjustment based on the result of machine scoring while mainly performing human evaluation.

  Next, the operation | movement of a human evaluation utilization scoring process is demonstrated with reference to the flowchart of FIG. When the singer gives an instruction to start the human evaluation use scoring process using the touch panel 15, the operation unit 25, or the remote controller 9, and the karaoke performance is started, the operation shown in FIG. 10 is performed.

  First, the CPU 11 of the karaoke apparatus 7 performs karaoke performance and starts singing voice mechanical scoring (s11). Next, the CPU 11 determines whether or not a predetermined section (for example, A melody) has elapsed (s12). If the predetermined section has elapsed, the CPU 11 aggregates the results of machine scoring for each predetermined section, and The machine scoring result and the singing voice in the section are transmitted to the server 1 (s13).

  When the CPU 51 of the server 1 receives the scoring result and the singing voice (s14), the CPU 51 performs the singing voice comparison process and extracts the pre-singing voice data having high similarity in the section (s15). Then, the average human scoring given to the extracted prior singing voice data is reflected in the scoring result of the current singing voice and output as the scoring result (s16). The scoring result is transmitted to the karaoke apparatus 7 (s17). Note that if there is no similar pre-song voice data, or only pre-song voice data with a low similarity (for example, 50% or less) exists, the karaoke apparatus 7 indicates “waiting for evaluation”. The information may be transmitted, and the karaoke apparatus 7 may display “Waiting for evaluation” or the like on the monitor 24 to display only the result of the machine scoring.

The singing voice transmitted from the karaoke apparatus 7 to the server 1 and the result of the machine scoring are accumulated as the pre-singing voice data in the HDD 53 of the server 1 and become the singing voice to be subject to human evaluation by the user. The CPU 11 receives the scoring result from the server 1 (s18), and displays the scoring result on the monitor 24 (s19). The scoring results preferably display the scoring results for each section and the overall scoring results obtained by averaging the sections up to now.

  Finally, the CPU 11 determines whether or not the music performance has ended (s20), and repeats the processing from s12 onward until the music performance ends. When the performance of the music is finished, the scoring result of the entire music may be displayed.

  In addition, in this embodiment, although the example which performs singing with the karaoke apparatus 7, transmits a singing voice | voice to the server 1, and performs a human evaluation utilization scoring process was shown, for example, a user's own PC and smart phone It is also possible to sing using an information processing device such as the server 1 or the information processing device to realize the singing voice evaluation device of the present invention. That is, the user sings using a PC or smartphone owned by the user, and the PC or smartphone transmits the singing voice to the server 1, thereby realizing the human evaluation use scoring process.

  In this case, the user instructs the start of the human evaluation use scoring process using the information processing apparatus owned by the user. When the user gives an instruction to start the human evaluation use scoring process, the information processing apparatus owned by the user performs a karaoke performance. The user's singing voice is transmitted to the server 1 (thereby, the CPU 51 of the server 1 realizes the singing voice input means of the present invention). Then, the CPU 51 of the server 1 performs a human evaluation use scoring process (s15 and subsequent processes in the flowchart shown in FIG. 5), thereby realizing the scoring means of the present invention. In this way, the singing voice can be transmitted from the information processing apparatus (or karaoke apparatus 7) owned by the user to the server 1, and the singing voice evaluation apparatus of the present invention can be realized by the server 1. It is also possible for the information processing device (or karaoke device 7) owned by the user to download the pre-song voice data from the server 1 and for the information processing device to perform the singing voice comparison process.

  In addition, in this embodiment, although the example which transmits song audio | voice (digital audio | voice signal) to the server 1 and accumulate | stores as prior song audio | voice data was shown, the image data (for example, a singer dancing while singing) was shown. Or the like) may be transmitted to the server 1 and stored as pre-singing voice data.

DESCRIPTION OF SYMBOLS 1 ... Server 2 ... Network 3 ... Karaoke store 4 ... Evaluator terminal 5 ... Relay machine 7 ... Karaoke apparatus 9 ... Remote control 11 ... CPU
12 ... RAM
13 ... HDD
14 ... Network I / F
DESCRIPTION OF SYMBOLS 15 ... Touch panel 16 ... Microphone 17 ... A / D converter 18 ... Sound source 19 ... Mixer 20 ... Sound system 21 ... Speaker 22 ... Decoder 23 ... Display processing part 24 ... Monitor 25 ... Operation part 26 ... Transmission / reception part

Claims (4)

Storage means for storing the singing voice and the human evaluation for the singing voice in advance and storing them as the pre-singing voice data;
Singing voice input means for inputting singing voice;
Scoring means for scoring the current singing voice input by the singing voice input means;
With
The scoring means compares the current singing voice and the pre-singing voice data, extracts pre-singing voice data similar to the current singing voice, and performs a human evaluation on the extracted pre-singing voice data. , Output in scoring results ,
The preliminary singing voice data includes a machine scoring result,
The scoring means performs a current singing machine scoring to compare the singing voice and the reference singing voice, and both the result of the current singing machine scoring and the result of the mechanical scoring included in the preliminary singing voice data. A singing voice evaluation device for outputting in the scoring results . The scoring means extracts a plurality of preliminary singing voice data similar to the current singing voice,
The singing voice evaluation apparatus according to claim 1, wherein each human evaluation is weighted according to the degree of similarity of each extracted prior singing voice data, and is included in the scoring result and output. The scoring means singing voice evaluation device according to claim 1 or claim 2 outputs the rating result by scoring the current singing voice for every predetermined section. A singing voice evaluation system comprising a server and a singer's terminal,
Storage means according to any one of claims 1 to 3 provided on one of the terminals or the server of the singer, the singer scoring means according to any one of claims 1 to 3 A singing voice evaluation system provided in either the terminal or the server, and the singing voice input means according to any one of claims 1 to 3 provided in the terminal of the singer.

Images