Performance recognition method and computer readable storage medium

Abstract

The application discloses a performance recognition method and a computer readable storage medium, and the method comprises the following steps: acquiring a sound played at a current moment as a current performance sound, and acquiring a sound corresponding to a score moment of the current moment in a score as a score sound corresponding to the current performance sound; calculating a same rate of the current performance sound and the score sound corresponding thereto; if the same rate of the current performance sound and the score sound corresponding thereto is less than a first threshold value, and there is a matching moment in n score moments adjacent to the score moment corresponding to the current moment, and a same rate of a sound of the matching moment and the current performance sound is greater than or equal to a second threshold value, then determining a best matching moment according to the same rate of the sound of each matching moment and the current performance sound; if the best matching moment is located before the score moment corresponding to the current moment, then determining that there is a rebound; and if the best matching moment is located after the score moment corresponding to the current moment, then determining that there is a bounce. The application can accurately detect whether a performer has a rebound or bounce behavior.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of performance recognition, in particular to a performance recognition method and a computer readable storage medium. BACKGROUND

[0002] Traditional musical instrument learning needs a special teacher to guide teaching, which is high in cost and low in efficiency.

[0003] At present, some musical instrument performance auxiliary software can be used to detect whether the user performs correctly, but the existing detection software can only detect according to the order of the notes in the music score. If the user has a performance error during the performance, wants to return to the wrong place to perform, or the notes at the current moment are missed by the user, and directly performs the notes of the next moment or the following measures, the existing software cannot normally identify the user's such back bounce or jump bounce behavior. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a performance recognition method and a computer readable storage medium, which can accurately detect whether the performer has a back bounce or jump bounce behavior.

[0005] In order to solve the above technical problems, the technical scheme adopted by the present application is as follows: a performance recognition method, comprising:

[0006] obtaining the note played at the current moment as the current performance note, and obtaining the music score moment corresponding to the current moment in the music score, taking the note of the music score moment corresponding to the current moment in the music score as the music score note corresponding to the current performance note, wherein the notes of the same music score moment in the music score represent one or more notes that need to be played at the same time;

[0007] calculating the same rate of the current performance note and the music score note corresponding thereto;

[0008] if the same rate of the current performance note and the music score note corresponding thereto is less than a preset first threshold value, and there is a matching moment in the n music score moments adjacent to the music score moment corresponding to the current moment, the same rate of the note of the matching moment and the current performance note is greater than or equal to a preset second threshold value, and n is a preset value, then determining the best matching moment according to the same rate of the note of each matching moment and the current performance note;

[0009] if the best matching moment is located before the music score moment corresponding to the current moment, it is determined as back bounce;

[0010] if the best matching moment is located after the music score moment corresponding to the current moment, it is determined as jump bounce.

[0011] The application further provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the method.

[0012] The application has the advantages that: by acquiring the sound played at the same time and matching it with the sound at the corresponding score time, the multiple sound recognition in a unit time can be realized; by calculating the same rate of the sound played at the same time and the sound at the corresponding score time, the playing condition at the time can be recognized; by finding the score time that best matches the sound played at the current time in the score and according to the positional relationship between the score time and the corresponding score time of the current time, it can be recognized whether there is a rebound behavior or a jump behavior, and the score time of rebound or the score time of jump can be obtained. The application detects and recognizes in time units, can accurately recognize the time position of rebound or jump behavior, and can also recognize the time position of rebound or jump, which is convenient for subsequent detection and recognition. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 The flowchart of the playing recognition method of the application is shown in Fig. 1.

[0014] Figure 2 The flowchart of the method of the first embodiment of the application is shown in Fig. 2.

[0015] Figure 3 The score diagram of the first embodiment of the application is shown in Fig. 3. DETAILED DESCRIPTION

[0016] In order to explain the technical content, the purposes and effects of the application in detail, the following will be described in detail in combination with the embodiments and the drawings.

[0017] Please refer to Figure 1 A playing recognition method, comprising:

[0018] Acquiring the sound played at the current time as the current playing sound, and acquiring the score time corresponding to the current time in the score, and taking the sound at the score time corresponding to the current time in the score as the score sound corresponding to the current playing sound, wherein the sound at the same score time in the score represents one or more sounds that need to be played at the same time;

[0019] Calculating the same rate of the current playing sound and the corresponding score sound;

[0020] If the same rate of the current playing sound and the corresponding score sound is less than a preset first threshold value, and there is a matching time in the n score times adjacent to the score time corresponding to the current time in the score, the same rate of the sound at the matching time and the current playing sound is greater than or equal to a preset second threshold value, and n is a preset value, then the best matching time is determined according to the same rate of the sound at each matching time and the current playing sound;

[0021] If the optimal matching time is before the time in the score corresponding to the current time, it is determined to be a rebound;

[0022] If the best matching time is after the time in the score corresponding to the current time, it is determined to be a bounce.

[0023] As can be seen from the above description, the beneficial effect of the present invention is that it can accurately detect whether the performer has a rebound or bounce behavior, and can obtain the time of the rebound or bounce of the score.

[0024] Furthermore, the calculation of the similarity rate between the currently played note and its corresponding musical score note specifically involves:

[0025] Match the pitch values ​​of each note in the currently played note with the pitch values ​​of each note in the corresponding musical score to determine the number of notes with the same pitch value;

[0026] Divide the number of notes with the same pitch value by the total number of notes in the corresponding musical score to obtain the similarity rate between the currently played note and its corresponding musical score note.

[0027] As described above, by calculating the same rate, it can be determined whether the performer is currently playing the note corresponding to the current moment in the score.

[0028] Furthermore, if the similarity rate between the currently played note and its corresponding note in the score is less than a preset first threshold, and there is a matching moment among the n adjacent notes in the score corresponding to the current moment, then the optimal matching moment is determined based on the similarity rate between the notes at each matching moment and the currently played note, specifically as follows:

[0029] Determine whether the similarity rate between the currently played note and its corresponding note in the musical score is less than a preset first threshold.

[0030] If not, the current time corresponding to the score will be taken as the best matching time, and any notes in the current performance that do not match the corresponding score notes will be marked as performance errors.

[0031] If so, determine whether there is a matching moment among the n musical moments that are adjacent to the musical moment corresponding to the current moment. The similarity rate between the note of the matching moment and the note played at the current moment is greater than or equal to a preset second threshold, where n is a preset value.

[0032] If it does not exist, the current time corresponding to the score is taken as the best matching time, and the notes in the current performance that do not match the corresponding score notes are marked as performance errors;

[0033] If a match exists, the optimal match time is determined based on the similarity rate between the notes at each match time and the currently played note.

[0034] From the above description, the higher the same rate of the current performance sound and the sound of a certain score moment, the more likely the performer is playing the sound of the score moment. During practice, the performer generally repeatedly practices a section and does not miss many score moments. Therefore, searching for matching moments near the score moment corresponding to the current moment can improve search efficiency and thus improve recognition efficiency.

[0035] Further, the best matching moment is determined according to the same rate of the sound of each matching moment and the current performance sound, specifically as follows.

[0036] If the number of matching moments is one, the matching moment is taken as the best matching moment.

[0037] If the number of matching moments is at least two, the matching moment with the highest same rate with the current performance sound is taken as the best matching moment.

[0038] From the above description, if the matching moment is unique, the only one matching moment is directly taken as the best matching moment. If the matching moment is not unique, the matching moment with the highest same rate is taken as the best matching moment.

[0039] Further, the matching moment with the highest same rate with the current performance sound is taken as the best matching moment, specifically as follows.

[0040] The matching moment with the highest same rate with the current performance sound is taken as the first matching moment, and it is determined whether the number of first matching moments is one.

[0041] If yes, the first matching moment is taken as the best matching moment.

[0042] If no, the best matching moment is determined according to the same rate of the sound of the next score moment of each first matching moment and the sound played at the next moment of the current moment.

[0043] From the above description, if there are multiple matching moments with the same and highest same rate, the performer is waited to play the sound of the next moment, and the same rate is compared with the sound of the next score moment of each to determine the best matching moment.

[0044] Further, the best matching moment is determined according to the same rate of the sound of the next score moment of each first matching moment and the sound played at the next moment of the current moment, specifically as follows.

[0045] The sound played at the next moment of the current moment is obtained as the next moment performance sound, and the same rate of the sound of the next score moment of each first matching moment and the next moment performance sound is calculated.

[0046] if the same rate of the note at the next score time and the note played at the next time is less than the preset second threshold value, the time interval between the score time corresponding to the current time and each first matching time is calculated respectively;

[0047] the first matching time with the least time interval between the score time corresponding to the current time is taken as the best matching time;

[0048] if the same rate of the note at the next score time and the note played at the next time of at least two first matching times is greater than or equal to the preset second threshold value, the first matching time with the highest same rate of the note at the next score time and the note played at the next time is taken as the best matching time.

[0049] From the above description, if the same rate of the note at the next score time and the note played at the next time of each first matching time does not exceed a certain value, the first matching time closest to the score time corresponding to the current time is taken as the best matching time, and if it exceeds, the first matching time with the highest same rate is taken as the best matching time.

[0050] Further, the first matching time with the least time interval between the score time corresponding to the current time is taken as the best matching time, specifically:

[0051] if the number of the first matching time with the least time interval between the score time corresponding to the current time is one, the first matching time is taken as the best matching time;

[0052] if the number of the first matching time with the least time interval between the score time corresponding to the current time is two, the first matching time after the score time corresponding to the current time among the two first matching times is taken as the best matching time.

[0053] Further, the first matching time with the highest same rate of the note at the next score time and the note played at the next time is taken as the best matching time, specifically:

[0054] the first matching time with the highest same rate of the note at the next score time and the note played at the next time is taken as the second matching time, and it is determined whether the number of the second matching time is one;

[0055] if yes, the second matching time is taken as the best matching time;

[0056] if no, the best matching time is determined according to the same rate of the note at the next score time and the note played at the second time after the current time of each second matching time.

[0057] From the above description, if the best matching time cannot be determined according to the sound of the next moment, the sound of the moment after the next moment is obtained to compare and judge, and the subsequent is carried out in this way.

[0058] Further, if the same rate of the current performance sound and the corresponding score sound is less than the first threshold, and there is a matching time in the n score moments adjacent to the score moment corresponding to the current moment, the best matching time is determined according to the same rate of the sound of each matching time and the current performance sound, and then the method further comprises:

[0059] The next score moment of the best matching time is taken as the score moment corresponding to the next moment of the current moment.

[0060] Further, the first threshold is less than or equal to the second threshold; the value range of the first threshold is (0, 1], and the value range of the second threshold is (0, 1].

[0061] The application also provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to realize the method.

[0062] Embodiment one

[0063] Please refer to Figures 2-3 The embodiment one of the application is a performance recognition method, which can be applied to a musical instrument performance auxiliary system.

[0064] Firstly, the following three definitions are given:

[0065] 1. Current performance sound, i.e. the sound played by the performer at the current moment; for the system, if the performance time between two sounds is less than a preset time threshold (which can be set according to the performance proficiency of the performer, such as 100 ms), the two sounds are considered to be played at the same moment.

[0066] 2. Score sound; the sound of the score moment corresponding to the current moment in the score is taken as the score sound corresponding to the current performance sound.

[0067] 3. Sound at the same score moment; the sound to be played simultaneously in the same time in the score represents the sound at the same score moment, such as Figure 3 , the notes in the same box in Figure 3 are the sound at the same score moment. In the performance process, the notes in the same box are played simultaneously to sound good.

[0068] 4. Score moment; the performance order of the sound in the score, the sound in the score is sorted according to the performance order, and the same serial number means that the sound needs to be played simultaneously.

[0069] 5. Repetition; for example, it is assumed that the current performance is to the third measure (i.e.Figure 3 The 6th musical moment in the measure marked 3, but the performer wants to go back to a few moments or measures earlier, such as the 2nd musical moment in the 3rd measure, or the 1st or 2nd measure. This is called a backspin.

[0070] 6. Skipping; For example, suppose the current piece is played at the first note of the second measure, but the performer wants to skip the second note of the second measure and play the third note of the second measure; or wants to skip several consecutive notes and start playing from the seventh note of the second measure; or wants to skip the second and third measures and start playing from the fourth measure. This is called skipping.

[0071] like Figure 2 As shown, the method in this embodiment includes the following steps:

[0072] S1: Obtain the note being played at the current moment as the current playing note, and obtain the musical note corresponding to the current moment in the score. Use the note of the musical note corresponding to the current moment in the score as the musical note corresponding to the current playing note.

[0073] One approach is to use multi-tone audio recognition to obtain the note being played at the current moment, or to use MIDI signal input. When the performer plays a single note or multiple notes within a unit of time, the performance signal is converted into a MIDI signal and transmitted to the recognition algorithm. MIDI signal transmission can transmit multiple MIDI signals within a unit of time.

[0074] Furthermore, in this embodiment, at the very beginning of the performance, the first musical moment in the score is taken as the musical moment corresponding to the first performance moment. After the best matching moment for the current moment is determined, the next musical moment of the best matching moment is taken as the musical moment corresponding to the next moment of the current moment.

[0075] In other words, for each moment during the performance, there is a corresponding musical moment in the score, which is the musical moment that the performer should play in the normal performance order. Then, using the method of this embodiment, the best matching moment is determined among the corresponding musical moment and the multiple musical moments on its left and right, which is the musical moment that the performer may actually play.

[0076] S2: Calculate the similarity rate between the currently played note and its corresponding note in the score.

[0077] Specifically, the pitch values ​​of each note in the currently played note are matched with the pitch values ​​of each note in the corresponding musical score to determine the number of notes with the same pitch value; then, the number of notes with the same pitch value is divided by the total number of notes in the corresponding musical score to obtain the similarity rate between the currently played note and its corresponding musical score.

[0078] S3: determining whether the same rate of the current performance sound and its corresponding score sound is less than a preset first threshold value, if yes, executing step S5, if no, executing step S4.

[0079] S4: taking the score time corresponding to the current time as the best matching time, and marking the sound corresponding to the current performance sound and not matching as a performance error, i.e. marking the sound with different pitch values as a performance error.

[0080] S5: determining whether there is a matching time in the n score times adjacent to the score time corresponding to the current time in the score, the sound of the matching time having a same rate with the current performance sound greater than or equal to a preset second threshold value, if yes, i.e. there is a matching time, executing step S6, if no, i.e. there is no matching time, executing step S4.

[0081] wherein n is a preset value. In the embodiment, n=50, i.e. searching in the 50 score times before and after the score time corresponding to the current time to determine whether there is a matching time in these adjacent score times. Further, if the number of score times before or after the score time corresponding to the current time is less than 50, all the score times before or after the score time are obtained.

[0082] S6: determining whether the number of the matching times is one, if yes, executing step S7, if no, indicating that the number of the matching times is at least two, executing step S8.

[0083] S7: taking the matching time as the best matching time. Then executing step S11.

[0084] S8: taking the matching time with the highest same rate with the current performance sound as the first matching time, and determining whether the number of the first matching time is one, if yes, executing step S9, if no, indicating that the number of the first matching time is at least two, i.e. indicating that the sounds of more than two matching times have the same rate with the current performance sound and the same rate is the highest, executing step S10.

[0085] S9: taking the first matching time as the best matching time. Then executing step S11.

[0086] S10: determining the best matching time according to the same rate of the sound of the next score time of each first matching time and the sound played at the next time of the current time. Then executing step S11.

[0087] Specifically, the sound played at the next time of the current time is obtained as the next time performance sound, and the same rate of the sound of the next score time of each first matching time and the next time performance sound is calculated respectively.

[0088] If the same rate of the note at the next score time of each first matching time and the note played at the next time is less than the preset second threshold, then the number of times between the score time corresponding to the current time and the score time corresponding to each first matching time is calculated respectively, and then the first matching time with the least number of times between the score time corresponding to the current time is taken as the best matching time, that is, the first matching time closest to the score time corresponding to the current time is taken as the best matching time.

[0089] Further, if the number of the first matching time with the least number of times between the score time corresponding to the current time is one, then the first matching time is directly taken as the best matching time; if the number of the first matching time with the least number of times between the score time corresponding to the current time is two, then the first matching time after the score time corresponding to the current time among the two first matching times is taken as the best matching time.

[0090] If only one first matching time has the same rate of the note at the next score time and the note played at the next time greater than or equal to the preset second threshold, then the first matching time is taken as the best matching time.

[0091] If more than two first matching times have the same rate of the note at the next score time and the note played at the next time greater than or equal to the preset second threshold, then the first matching time with the highest same rate of the note at the next score time and the note played at the next time is taken as the best matching time.

[0092] Further, the first matching time with the highest same rate of the note at the next score time and the note played at the next time is taken as the second matching time, if the number of the second matching time is one, then the second matching time is directly taken as the best matching time; if the number of the second matching time is more than two, then the best matching time is determined according to the same rate of the note at the next score time of each second matching time and the note played at the second time after the current time, and the specific implementation process can be referred to the above description.

[0093] That is, if the best matching time cannot be determined according to the note at the next time, then the note at the time after the next time is obtained to compare and judge, and the subsequent is carried out in this way.

[0094] S11: Determine whether the best matching time is before the score time corresponding to the current time, if yes, execute step S12, if not, it means that the best matching time is after the score time corresponding to the current time, then execute step S13.

[0095] S12: Determine that it is rebound, that is, it is considered that the performer has rebound behavior.

[0096] S13: Determine that the bounce occurs, that is, consider that the performer has a bounce behavior.

[0097] For example, assuming that the best matching time of the last time is the first music time of the second measure in the music score, the music time corresponding to the current time is the second music time of the second measure, at this time, if the best matching time is identified as the first music time of the first measure, it is considered that the performer has a bounce; if the best matching time is identified as the first music time of the third measure, it is considered that the performer has a bounce.

[0098] Further, the first threshold is less than or equal to the second threshold. In the embodiment, the value range of the first threshold is (0, 1], and the value range of the second threshold is (0, 1].

[0099] The embodiment detects and identifies in time units, can accurately identify the time position of the bounce or bounce behavior, and can also identify the time position of the bounce or the time position of the bounce, which is convenient for subsequent detection and identification.

[0100] Embodiment two

[0101] The embodiment is a computer readable storage medium corresponding to the above embodiment, which stores a computer program. The program is executed by a processor to realize the steps of the above-mentioned embodiment, and the same technical effect can be achieved, which is not repeated here.

[0102] In summary, the present application provides a kind of performance identification method and computer readable storage medium, by obtaining the sound played at the same time, and it is matched with the sound of corresponding music time, multiple sound identification in unit time can be realized;By calculating the same rate of the sound played at the same time and the sound of corresponding music time, the performance at the time can be identified;By finding the music time most matched with the sound played at the current time in the music score, and according to the position relationship between it and the music time corresponding to the current time, whether there is bounce behavior or bounce behavior can be identified, and the bounce music time or bounce music time can be obtained.

[0103] The above is only an embodiment of the present application, and does not limit the patent scope of the present application, any equivalent transformation or direct or indirect application in related technical field based on the content of the specification and drawings of the present application is also included in the patent protection scope of the present application.

Claims

1. A performance recognition method characterized by comprising: The method comprises the following steps: acquiring a sound played at a current time as a current playing sound, and acquiring a score time corresponding to the current time in a score, and acquiring a sound at the score time corresponding to the current time in the score as a score sound corresponding to the current playing sound, wherein the sounds at the same score time in the score represent one or more sounds that need to be played simultaneously; calculating a same rate of the current playing sound and the corresponding score sound; if the same rate of the current playing sound and the corresponding score sound is less than a preset first threshold value, and there is a matching time in n score times adjacent to the score time corresponding to the current time in the score, the same rate of the sound at the matching time and the current playing sound is greater than or equal to a preset second threshold value, and n is a preset value, then determining a best matching time according to the same rate of the sound at each matching time and the current playing sound; if the best matching time is located before the score time corresponding to the current time, then determining that it is a rebound; if the best matching time is located after the score time corresponding to the current time, then determining that it is a bounce; if the same rate of the current playing sound and the corresponding score sound is less than the preset first threshold value, and there is a matching time in n score times adjacent to the score time corresponding to the current time in the score, the same rate of the sound at the matching time and the current playing sound is greater than or equal to the preset second threshold value, and n is a preset value, then determining the best matching time according to the same rate of the sound at each matching time and the current playing sound, specifically as follows: determining whether the same rate of the current playing sound and the corresponding score sound is less than the preset first threshold value; if not, then taking the score time corresponding to the current time as the best matching time, and marking the sound in the current playing sound that does not match the corresponding score sound as a playing error; if yes, then determining whether there is a matching time in n score times adjacent to the score time corresponding to the current time in the score, the same rate of the sound at the matching time and the current playing sound is greater than or equal to the preset second threshold value, and n is a preset value; if not, then taking the score time corresponding to the current time as the best matching time, and marking the sound in the current playing sound that does not match the corresponding score sound as a playing error; if yes, then determining the best matching time according to the same rate of the sound at each matching time and the current playing sound.

2. The performance recognition method according to claim 1, characterized by, The calculation of the same rate of the current playing sound and the corresponding score sound is specifically as follows: matching the pitch values of the sounds in the current playing sound with the pitch values of the sounds in the corresponding score sound to determine the number of sounds with the same pitch values; dividing the number of sounds with the same pitch values by the total number of sounds in the corresponding score sound to obtain the same rate of the current playing sound and the corresponding score sound.

3. The performance recognition method according to claim 1, characterized by, The determination of the best matching time according to the same rate of the sound at each matching time and the current playing sound is specifically as follows: if the number of matching times is one, then taking the matching time as the best matching time; if the number of matching times is at least two, then taking the matching time with the highest same rate with the current playing sound as the best matching time.

4. The performance recognition method according to claim 3, characterized by, The taking of the matching time with the highest same rate with the current playing sound as the best matching time is specifically as follows: taking the matching time with the highest same rate with the current playing sound as a first matching time, and determining whether the number of first matching times is one; if not, then taking the score time corresponding to the current time as the best matching time, and marking the sound in the current playing sound that does not match the corresponding score sound as a playing error; if yes, then determining whether there is a matching time in n score times adjacent to the score time corresponding to the current time in the score, the same rate of the sound at the matching time and the current playing sound is greater than or equal to the preset second threshold value, and n is a preset value; if not, then taking the score time corresponding to the current time as the best matching time, and marking the sound in the current playing sound that does not match the corresponding score sound as a playing error; if yes, then determining the best matching time according to the same rate of the sound at each matching time and the current playing sound. If yes, the first matching time is taken as the best matching time. If no, the best matching time is determined according to the same rate of the note at the next score time of each first matching time and the note played at the next time of the current time.

5. The performance recognition method according to claim 4, characterized by, The best matching time is determined according to the same rate of the note at the next score time of each first matching time and the note played at the next time of the current time, specifically as follows: The note played at the next time of the current time is taken as the next time playing note, and the same rate of the note at the next score time of each first matching time and the next time playing note is calculated respectively. If the same rate of the note at the next score time of each first matching time and the next time playing note is less than a preset second threshold, the number of time intervals between the score time corresponding to each first matching time and the score time corresponding to the current time is calculated respectively. The first matching time with the least number of time intervals between the score time corresponding to the current time is taken as the best matching time. If the same rate of the note at the next score time of at least two first matching times and the next time playing note is greater than or equal to the preset second threshold, the first matching time with the highest same rate of the note at the next score time and the next time playing note is taken as the best matching time.

6. The performance recognition method according to claim 5, characterized by, The first matching time with the least number of time intervals between the score time corresponding to the current time is taken as the best matching time, specifically as follows: If the number of first matching times with the least number of time intervals between the score time corresponding to the current time is one, the one first matching time is taken as the best matching time. If the number of first matching times with the least number of time intervals between the score time corresponding to the current time is two, the first matching time located after the score time corresponding to the current time among the two first matching times is taken as the best matching time.

7. The performance recognition method according to claim 5, characterized by, The first matching time with the highest same rate of the note at the next score time and the next time playing note is taken as the best matching time, specifically as follows: The first matching time with the highest same rate of the note at the next score time and the next time playing note is taken as the second matching time, and it is determined whether the number of second matching times is one. If yes, the second matching time is taken as the best matching time. If no, the best matching time is determined according to the same rate of the note at the next score time of each second matching time and the note played at the second time after the current time.

8. The performance recognition method according to any one of claims 1 to 2, 4 to 7, characterized in that, After the best matching time is determined according to the same rate of the note at the next score time of each matching time and the current playing note, and if the same rate of the note at the next score time of the best matching time and the next time corresponding to the current time is less than a preset first threshold, and there is a matching time in n score times adjacent to the score time corresponding to the current time in the score, the next score time of the best matching time is taken as the score time corresponding to the next time of the current time. The first threshold is less than or equal to the second threshold; the value range of the first threshold is (0, 1], and the value range of the second threshold is (0, 1].

9. The performance recognition method according to any one of claims 1-2, 4-7, characterized by, The program is executed by the processor to realize the method of any one of claims 1-9.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to realize the method of any one of claims 1-9.

Images