WO2020015411A1 - 一种训练改编水平评价模型、评价改编水平的方法及装置 - Google Patents

一种训练改编水平评价模型、评价改编水平的方法及装置 Download PDF

Info

Publication number
WO2020015411A1
WO2020015411A1 PCT/CN2019/083092 CN2019083092W WO2020015411A1 WO 2020015411 A1 WO2020015411 A1 WO 2020015411A1 CN 2019083092 W CN2019083092 W CN 2019083092W WO 2020015411 A1 WO2020015411 A1 WO 2020015411A1
Authority
WO
WIPO (PCT)
Prior art keywords
sample
musical
track
work
adapted musical
Prior art date
Application number
PCT/CN2019/083092
Other languages
English (en)
French (fr)
Inventor
杨新颖
Original Assignee
阿里巴巴集团控股有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 阿里巴巴集团控股有限公司 filed Critical 阿里巴巴集团控股有限公司
Priority to EP19838690.6A priority Critical patent/EP3779814A4/en
Priority to SG11202010627QA priority patent/SG11202010627QA/en
Publication of WO2020015411A1 publication Critical patent/WO2020015411A1/zh
Priority to US17/083,740 priority patent/US11074897B2/en
Priority to US17/382,860 priority patent/US11367424B2/en

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
    • G10H1/00—Details of electrophonic musical instruments
    • G10H1/0008—Associated control or indicating means
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G06Q50/205—Education administration or guidance
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
    • G10H2210/00—Aspects or methods of musical processing having intrinsic musical character, i.e. involving musical theory or musical parameters or relying on musical knowledge, as applied in electrophonic musical tools or instruments
    • G10H2210/031—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal
    • G10H2210/091—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal for performance evaluation, i.e. judging, grading or scoring the musical qualities or faithfulness of a performance, e.g. with respect to pitch, tempo or other timings of a reference performance

Definitions

  • the embodiments of the present specification relate to the field of information technology, and in particular, to a training adaptation level evaluation model and a method and device for evaluating the adaptation level.
  • Music adaptation refers to re-arrangement of tunes based on existing music works as the material to obtain adapted music works.
  • the evaluation of the level of adaptation of musical works mainly depends on manual labor, and it is difficult to efficiently process a large number of adapted music works.
  • the embodiments of the present specification provide a training adaptation evaluation model, a method and a device for evaluating the adaptation level, and the technical scheme is as follows:
  • a method for training an adapted level evaluation model including:
  • each adapted musical composition sample in the adapted musical composition set For each adapted musical composition sample in the adapted musical composition set, separate each audio track of the adapted musical composition sample, and determine information of the playing instrument corresponding to each of the adapted musical composition samples, respectively;
  • a combination of the performance instrument information corresponding to the track and the originality characteristic value of the musical tone is used as a sample feature of the sample of the adapted musical work;
  • Each sample feature of the adapted musical work sample is used as a model input, and the known adapted score of the adapted musical work sample is used as a model output, and model training is performed to obtain an adapted level evaluation model.
  • a method for evaluating an adaptation level including:
  • a combination of the performance instrument information corresponding to the audio track and the musical original characterization value is used as a feature of the adapted musical work;
  • the adapted level evaluation model is obtained according to the method for training and adapted the level evaluation model.
  • an apparatus for training an adapted level evaluation model including:
  • An acquisition module to acquire a sample collection of adapted musical works
  • An audio track separation module for each adapted music work sample in the adapted music work set, separates each track of the adapted music work sample, and determines the performance instrument information corresponding to each track of the adapted music work sample. ;
  • the originality analysis module performs musical originality analysis on each track of the adapted musical work sample, and obtains musical originality characteristic values corresponding to each track of the adapted musical work sample;
  • a determination module for each audio track of the adapted musical work sample, as a sample feature of the adapted musical work sample, a combination of the performance instrument information corresponding to the audio track and a musical original characteristic value;
  • the training module takes each sample feature of the adapted musical work sample as a model input, and uses a known adapted score of the adapted music work sample as a model output, and performs model training to obtain an adapted level evaluation model.
  • a device for evaluating an adaptation level including:
  • An audio track separation module which separates each audio track of the adapted musical composition, and determines information of a playing instrument corresponding to each audio track of the adapted musical composition
  • the originality analysis module performs music originality analysis on each track of the adapted music work, and obtains the originality characteristic values of the music sound corresponding to each track of the adapted music work;
  • a determining module for each audio track of the adapted musical composition, as a feature of the adapted musical composition, a combination of the performance instrument information corresponding to the audio track and a musical original characteristic value;
  • An evaluation module inputting each feature of the adapted musical work into an adapted level evaluation model to obtain an adapted level score output by the adapted level evaluation model corresponding to the adapted musical work;
  • the adapted level evaluation model is obtained according to the above-mentioned method of training and adapted level evaluation model.
  • the technical solution provided in the embodiment of the present specification uses the method of machine learning to train an adaptation level evaluation model based on the sample characteristics of the adapted music composition samples of a plurality of known adaptation level scores for evaluating the adapted music composition to be evaluated.
  • the sample characteristics of the adapted musical work sample include at least a combination of the performance instrument information and musical original characteristic values corresponding to each track of the adapted musical work sample.
  • FIG. 1 is a schematic flowchart of a method for training an adapted level evaluation model according to an embodiment of the present specification
  • FIG. 2 is a schematic flowchart of a method for evaluating an adaptation level provided by an embodiment of the present specification
  • FIG. 3 is a schematic structural diagram of an apparatus for training an adapted level evaluation model provided by an embodiment of the present specification
  • FIG. 4 is a schematic structural diagram of an apparatus for evaluating an adaptation level provided by an embodiment of the present specification
  • FIG. 5 is a schematic structural diagram of a computer device for configuring a method according to an embodiment of the present specification.
  • FIG. 1 is a schematic flowchart of a method for training and adapting a level evaluation model according to an embodiment of the present specification, including the following steps:
  • the execution subject of the method may be a device having a data processing function, such as a server, a computer, a mobile phone, and the like.
  • MIDI Musical Instrument Digital Interface
  • S104 Perform a musical originality analysis on each track of the adapted musical work sample to obtain a musical originality characteristic value corresponding to each track of the adapted musical work sample.
  • model training As we all know, in the field of model training, you can extract features from samples, and for each sample, use a supervised learning algorithm to perform model training based on the sample features of the sample and the label of the sample.
  • the process of model training is actually a process in which the machine learns the mapping relationship between the sample features of the samples and the labels. Therefore, after the model is trained, the features of the target object to be verified can be input to the model, and the model outputs the label of the target object.
  • steps S102 to S106 may be performed.
  • the melody corresponding to each track is performed by one or more playing instruments, and the corresponding playing instruments of each track are different.
  • the adapted musical work sample is a MIDI format arranger file
  • the melody of the MIDI format arranger file is The audio tracks are arranged separately, and the sample of the adapted musical works also includes information of the playing instrument corresponding to each track, so it is easy to separate the individual tracks of the sample of the adapted musical works and obtain the sample of the adapted musical works
  • Each track corresponds to the performance instrument information.
  • the performance instrument information is generally a preset performance instrument number. For example, the guitar performance instrument number is 1 and the bass performance instrument number is 5.
  • the performance instrument information may also be another character string that can uniquely identify the performance instrument.
  • the sample of the adapted musical work is an arrangement file in a digital format other than the MIDI format
  • common track separation techniques can be used to separate the individual tracks of the adapted musical work sample, and then according to Different tonal characteristics of various playing instruments, analyze the playing instruments corresponding to each track of the adapted musical work sample.
  • a performance instrument number can be assigned to each performance instrument in advance. In this way, the performance instrument corresponding to each track of the adapted musical composition sample is analyzed, and the adapted musical composition sample is determined. Each track corresponds to the performance instrument information.
  • musical originality analysis refers to the comparison of the target melody with the reference melody in the melody library based on the existing melody library using digital analysis methods to determine the target melody from the melody library.
  • the most similar reference melody is used as the key reference melody, and the similarity between the target melody and the key reference melody is obtained. This is a technical means well known to those skilled in the art.
  • the similarity is generally a value between 0 and 100%, which represents the degree of similarity between the target melody and the key reference melody. For example, the similarity is 20%, indicating that the target melody is 20% similar to the key reference melody. Obviously, the higher the similarity, the lower the originality of the target melody.
  • each track of the adapted musical work sample corresponds to a piece of melody. Therefore, each of the adapted musical works sample Musical originality analysis is performed on each track, and the original musicality representative value corresponding to each audio track of the adapted musical composition sample can be obtained.
  • the originality characteristic value of the musical sound corresponding to the soundtrack may be the similarity between the melody corresponding to the soundtrack and the key reference melody in the melody library. At this time, the larger the originality characteristic value of the musical sound , The less original the melody corresponding to that track.
  • the musical sound originality characteristic value corresponding to the audio track may also be (1-the similarity). At this time, the larger the musical sound originality characteristic value, the greater the originality of the melody corresponding to the music track.
  • the musical sound originality characteristic value corresponding to the audio track may also be obtained by performing other processing on the similarity.
  • the originality characteristic value of the music corresponding to the audio track can represent the originality of the melody corresponding to the audio track.
  • a sample of any adapted musical work in the adapted musical works collection is taken as an example to explain, for each track of the adapted musical work sample, the performance instrument information and musical originals corresponding to the track
  • the combination of sexual characterization values serves as a sample feature of the adapted musical composition sample.
  • the playing instrument number corresponding to the audio track A is further determined as 1 (guitar).
  • the performance instrument number corresponding to track B is 5 (bass), and the performance instrument number corresponding to track C is 3 (drum).
  • a musical originality analysis is performed on each track of the adapted music sample, and the musical originality characteristic value corresponding to track A is 70, the musical originality characteristic value corresponding to track B is 50, and the audio track C corresponds to Has a musical originality of 30.
  • the combination of the three playing instrument numbers (1, 70), (5, 50), and (3, 30) and the originality characteristic values of the musical tone can be used as three samples of the adapted musical work sample. feature.
  • the adaptation level score corresponding to the sample of the adapted musical work is manually specified.
  • the adaptation level of the adapted musical work sample may be evaluated by several music experts based on subjective experience, and a score of the adapted level of the adapted musical work sample may be given.
  • each sample feature of the adapted musical work sample may be used as a model input, and The score of the adaptation level corresponding to the sample of the adapted musical work is model output, and model training is performed. Finally, an adapted level evaluation model was obtained.
  • step S102 for a certain track of the adapted musical works sample, the following situations may occur:
  • the playing instrument corresponding to this track may be a relatively rare playing instrument.
  • a designated number can be preset as the unknown playing instrument number for uniformly identifying the unrecognizable playing instrument and the rare playing instrument.
  • the performance instrument information corresponding to the audio track is sometimes a preset performance instrument number, or sometimes a designated number.
  • the designated label may be 99.
  • the playing instrument number corresponding to track A is determined to be 99; for the adaptation
  • the playing instrument number corresponding to the track B is also determined to be 99.
  • the playing instrument corresponding to the track C is a drum. Among the preset playing instruments, the playing instrument number corresponding to the track C is determined as the drum drum corresponding number. , Such as 3.
  • the sample features of the adapted musical work sample may include not only a combination of information of a number of playing musical instruments and original characterization values of musical sounds, but also at least one of the following features:
  • the number of pre-designated playing instrument information Specifically, before performing model training, for each adapted musical work sample, determine the number of pre-designated musical instrument information among the performing musical instrument information corresponding to the adapted musical work sample; and use the determined quantity as the adapted musical work sample corresponding to A sample feature.
  • Coordination characterization values corresponding to samples of adapted musical works Specifically, before performing model training, for each adapted musical work sample, perform a coordinated analysis on each track of the adapted musical work sample to obtain a coordinated representative value corresponding to the adapted musical work sample; and the obtained coordinatedness The representative value is used as a sample feature corresponding to the adapted musical composition sample.
  • the coordination analysis of each track is to analyze whether the rhythm of the corresponding melody of each track is in sync, and the degree of tempo of the corresponding melody of each track of the adapted music sample is to use the coordination of the corresponding sample of the adapted music work. Characterization value to characterize.
  • the performance instrument information designated in advance is 1, 5, and 6.
  • the performance number corresponding to track A of the adapted musical work sample is 1 (guitar)
  • the performance number corresponding to track B is 5 (bass)
  • the performance number corresponding to track C is 99 (unknown playing instrument). It can be seen that, among the performance instrument information corresponding to the adapted musical composition sample, the number of performance instrument information (ie, number 1 and number 5) specified in advance is two.
  • the musical originality characteristic value of track A is 20
  • the musical originality characteristic value of track B is 60
  • the musical originality characteristic value of track C is 80
  • the coordinated representative value of the adapted musical composition sample is 59.
  • the sample features of the adapted musical composition sample are (1, 20), (5, 60), (99, 80), 2, 59, a total of 5 sample features.
  • FIG. 2 is a schematic flowchart of a method for evaluating an adaptation level provided by an embodiment of the present specification, including the following steps:
  • S202 Separate each audio track of the adapted music composition, and determine information of the playing instrument corresponding to each audio track of the adapted music composition.
  • S204 Perform a musical originality analysis on each track of the adapted musical composition to obtain a musical originality characteristic value corresponding to each of the audio tracks of the adapted musical composition.
  • the method for evaluating an adaptation level shown in FIG. 2 is actually a method of evaluating an adapted musical work to be evaluated by using an adaptation level evaluation model obtained by training an adaptation level evaluation model shown in FIG. 1.
  • the adapted level evaluation model will output the adapted corresponding to the adapted musical works.
  • Level score Generally, the higher the adaptation level score, the higher the adaptation level.
  • the performance instrument information corresponding to the audio track is corresponding to the number of a known performance instrument, or corresponding to the number of an unknown performance instrument.
  • the amount of pre-designated musical instrument information in the performance musical instrument information corresponding to the adapted musical work may be determined; and the determined amount is used as the adapted music A feature corresponding to the work.
  • a coordinated analysis may be performed on each track of the adapted musical work to obtain a coordinated representative value corresponding to the adapted musical work;
  • the coordinated representative value of s is used as a feature corresponding to the adapted musical composition.
  • the trained adaptation level evaluation model can efficiently complete the evaluation of the adaptation level of a large number of adapted musical works.
  • performing model training on the characteristics of the musical instrument information and musical original characterization values corresponding to each track of the adapted musical composition can make the adapted level evaluation model more accurate in evaluating the adapted musical composition.
  • the embodiment of the present specification also provides a device for training and adapting the level evaluation model.
  • the device includes:
  • An acquisition module 301 which obtains a sample set of adapted musical works
  • An audio track separation module 302 for each adapted musical work sample in the adapted musical work set, separates each track of the adapted musical work sample, and determines the playing instrument corresponding to each track of the adapted musical work sample information;
  • the originality analysis module 303 performs music originality analysis on each track of the adapted music work sample to obtain a musical originality characteristic value corresponding to each track of the adapted music work sample;
  • the determining module 304 for each audio track of the adapted musical work sample, uses a combination of the performance instrument information corresponding to the audio track and the musical original characterization value as a sample feature of the adapted musical work sample;
  • the training module 305 uses each sample feature of the adapted musical work sample as a model input, and uses a known adapted score of the adapted music work sample as a model output to perform model training to obtain an adapted level evaluation model.
  • the corresponding performance instrument information of the track is a preset performance instrument number, or a designated number.
  • the training module 305 determines, before performing model training, the number of pre-designated performance instrument information in the performance instrument information corresponding to the adapted musical composition sample; and using the determined quantity as a sample feature corresponding to the adapted musical composition sample.
  • the training module 305 performs coordination analysis on each track of the adapted musical composition sample before performing model training to obtain a coordinated representation value corresponding to the adapted musical composition sample; and uses the obtained coordinated representation value as the adaptation A sample feature corresponding to a sample of a musical work.
  • the adapted level score corresponding to the adapted musical work sample is manually specified.
  • the embodiment of the present specification also correspondingly provides a device for evaluating the adaptation level.
  • the device includes:
  • An acquisition module 401 which acquires an adapted musical work
  • An audio track separation module 402 which separates each audio track of the adapted musical work, and determines information of playing musical instruments corresponding to each audio track of the adapted musical work;
  • the originality analysis module 403 performs music originality analysis on each track of the adapted music piece to obtain the originality characteristic value of the music piece corresponding to each track of the adapted music piece;
  • a determining module 404 for each audio track of the adapted musical composition, as a feature of the adapted musical composition, a combination of the performance instrument information and musical original characteristic values corresponding to the audio segment;
  • the adapted level evaluation model is obtained according to the method for training and adapted the level evaluation model shown in FIG. 1.
  • the performance instrument information corresponding to the track is corresponding to the number of a known performance instrument, or corresponding to the number of an unknown performance instrument.
  • the evaluation module 405 determines the number of pre-designated musical instrument information in the performance instrument information corresponding to the adapted music composition before inputting each feature of the adapted music composition into the adaptation level evaluation model; the determined number As a feature corresponding to the adapted musical composition.
  • the evaluation module 405 performs a coordination analysis on each audio track of the adapted musical composition before inputting the sample characteristics of the adapted musical composition into the adapted level evaluation model to obtain the corresponding coordination of the adapted musical composition. Characteristic value; the obtained coordinated characteristic value is taken as a feature corresponding to the adapted musical composition.
  • An embodiment of the present specification further provides a computer device including at least a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the programs shown in FIG. 1 and FIG. 2 when the program is executed. The function of the method.
  • FIG. 5 shows a more specific schematic diagram of the hardware structure of a computing device provided by an embodiment of the present specification.
  • the device may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050.
  • the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 implement a communication connection within the device through a bus 1050.
  • the processor 1010 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits, etc., for performing related operations. Program to implement the technical solutions provided by the embodiments of this specification.
  • a general-purpose CPU Central Processing Unit
  • ASIC Application Specific Integrated Circuit
  • the memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage devices, dynamic storage devices, and the like.
  • the memory 1020 may store an operating system and other application programs.
  • related program codes are stored in the memory 1020 and are called and executed by the processor 1010.
  • the input / output interface 1030 is used to connect an input / output module to implement information input and output.
  • the input / output / module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions.
  • the input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc.
  • the output device may include a display, a speaker, a vibrator, and an indicator light.
  • the communication interface 1040 is used to connect a communication module (not shown in the figure) to implement communication interaction between the device and other devices.
  • the communication module can implement communication through a wired method (such as USB, network cable, etc.), and can also implement communication through a wireless method (such as mobile network, WIFI, Bluetooth, etc.).
  • the bus 1050 includes a path for transmitting information between various components of the device (for example, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).
  • the device may also include necessary for achieving normal operation Other components.
  • the foregoing device may also include only components necessary to implement the solutions of the embodiments of the present specification, and does not necessarily include all the components shown in the drawings.
  • An embodiment of the present specification also provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the functions of the methods described in FIG. 1 and FIG. 2 are implemented.
  • Computer-readable media includes both permanent and non-persistent, removable and non-removable media.
  • Information can be stored by any method or technology.
  • Information may be computer-readable instructions, data structures, modules of a program, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), and read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, read-only disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media may be used to store information that can be accessed by computing devices.
  • computer-readable media does not include temporary computer-readable media, such as modulated data signals and carrier waves.
  • the embodiments of the present specification can be implemented by means of software plus a necessary universal hardware platform. Based on such an understanding, the technical solutions of the embodiments of the present specification may be embodied in the form of software products that are essentially or contribute to the existing technology.
  • the computer software product may be stored in a storage medium, such as ROM / RAM, Magnetic disks, optical disks, and the like include a number of instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the methods described in various embodiments or portions of the embodiments of this specification.
  • the system, method, module, or unit described in the foregoing embodiments may be specifically implemented by a computer chip or entity or a product with a certain function.
  • a typical implementation device is a computer, and the specific form of the computer may be a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email sending and receiving device, and a game control Desk, tablet computer, wearable device, or a combination of any of these devices.
  • each embodiment in this specification is described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.
  • the description is relatively simple, and for the relevant parts, refer to the description of the method embodiments.
  • the method embodiments described above are only schematic, and the modules described as separate components may or may not be physically separated. When implementing the solutions of the embodiments of this specification, the functions of each module may be the same. Or multiple software and / or hardware. Some or all of the modules may also be selected according to actual needs to achieve the objective of the solution of this embodiment. Those of ordinary skill in the art can understand and implement without creative efforts.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Educational Administration (AREA)
  • Physics & Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Development Economics (AREA)
  • Marketing (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Game Theory and Decision Science (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Educational Technology (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Auxiliary Devices For Music (AREA)

Abstract

一种训练改编水平评价模型、评价改编水平的方法及装置。根据多个已知改编水平评分的改编音乐作品样本的样本特征，采用机器学***评价模型，用于对待评价的改编音乐作品进行评价。其中，改编音乐作品样本的样本特征至少包括改编音乐作品样本的每个音轨对应的演奏乐器信息与乐音原创性表征值的组合。

Description

一种训练改编水平评价模型、评价改编水平的方法及装置 技术领域
本说明书实施例涉及信息技术领域，尤其涉及一种训练改编水平评价模型、评价改编水平的方法及装置。
背景技术
音乐改编，是指以现有的音乐作品为素材，重新进行曲调编配工作，得到改编音乐作品。
随着线上音乐市场的繁荣，大量的改编音乐作品被上传到互联网。在有些业务场景下，需要对改编音乐作品的改编水平进行评价。例如，版权业务平台需要对用户上传的改编音乐作品的改编水平进行评价，以便依据评价结果决定应向该用户发放多少奖励。
目前，评价改编音乐作品的改编水平主要依赖人工，难以高效处理大量的改编音乐作品。
发明内容
为了解决现有的评价改编音乐作品的改编水平的方式效率较低的问题，本说明书实施例提供一种训练改编水平评价模型、评价改编水平的方法及装置，技术方案如下：
根据本说明书实施例的第1方面，提供一种训练改编水平评价模型的方法，包括：
获取改编音乐作品样本集合；
针对所述改编音乐作品集合中的每个改编音乐作品样本，分离出该改编音乐作品样本的各音轨，并确定该改编音乐作品样本的各音轨分别对应的演奏乐器信息；
分别对该改编音乐作品样本的各音轨进行乐音原创性分析，得到该改编音乐作品样本的各音轨分别对应的乐音原创性表征值；
针对该改编音乐作品样本的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为该改编音乐作品样本的一个样本特征；
以该改编音乐作品样本的各样本特征为模型输入，以已知的，该改编音乐作品样本对应的改编水平评分为模型输出，进行模型训练，以便得到改编水平评价模型。
根据本说明书实施例的第2方面，提供一种评价改编水平的方法，包括：
获取改编音乐作品；
分离出所述改编音乐作品的各音轨，并确定所述改编音乐作品的各音轨分别对应的演奏乐器信息；
分别对所述改编音乐作品的各音轨进行乐音原创性分析，得到所述改编音乐作品的各音轨分别对应的乐音原创性表征值；
针对所述改编音乐作品的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为所述改编音乐作品的一个特征；
将所述改编音乐作品的各特征输入到改编水平评价模型，以得到所述改编水平评价模型输出的对应于所述改编音乐作品的改编水平评分；
其中，所述改编水平评价模型是根据上述训练改编水平评价模型的方法得到的。
根据本说明书实施例的第3方面，提供一种训练改编水平评价模型的装置，包括：
获取模块，获取改编音乐作品样本集合；
音轨分离模块，针对所述改编音乐作品集合中的每个改编音乐作品样本，分离出该改编音乐作品样本的各音轨，并确定该改编音乐作品样本的各音轨分别对应的演奏乐器信息；
原创性分析模块，分别对该改编音乐作品样本的各音轨进行乐音原创性分析，得到该改编音乐作品样本的各音轨分别对应的乐音原创性表征值；
确定模块，针对该改编音乐作品样本的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为该改编音乐作品样本的一个样本特征；
训练模块，以该改编音乐作品样本的各样本特征为模型输入，以已知的，该改编音乐作品样本对应的改编水平评分为模型输出，进行模型训练，以便得到改编水平评价模型。
根据本说明书实施例的第4方面，提供一种评价改编水平的装置，包括：
获取模块，获取改编音乐作品；
音轨分离模块，分离出所述改编音乐作品的各音轨，并确定所述改编音乐作品的各音轨分别对应的演奏乐器信息；
原创性分析模块，分别对所述改编音乐作品的各音轨进行乐音原创性分析，得到所述改编音乐作品的各音轨分别对应的乐音原创性表征值；
确定模块，针对所述改编音乐作品的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为所述改编音乐作品的一个特征；
评价模块，将所述改编音乐作品的各特征输入到改编水平评价模型，以得到所述改编水平评价模型输出的对应于所述改编音乐作品的改编水平评分；
其中，所述改编水平评价模型是根据上述的训练改编水平评价模型的方法得到的。
本说明书实施例所提供的技术方案，根据多个已知改编水平评分的改编音乐作品样本的样本特征，采用机器学***评价模型，用于对待评价的改编音乐作品进行评价。其中，改编音乐作品样本的样本特征至少包括改编音乐作品样本的每个音轨对应的演奏乐器信息与乐音原创性表征值的组合。通过本说明书实施例，利用训练好的改编水平评价模型，可以高效地完成对大量改编音乐作品的改编水平评价工作。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本说明书实施例。
此外，本说明书实施例中的任一实施例并不需要达到上述的全部效果。
附图说明
为了更清楚地说明本说明书实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本说明书实施例中记载的一些实施例，对于本领域普通技术人员来讲，还可以根据这些附图获得其他的附图。
图1是本说明书实施例提供的一种训练改编水平评价模型的方法的流程示意图；
图2是本说明书实施例提供的一种评价改编水平的方法的流程示意图；
图3是本说明书实施例提供的一种训练改编水平评价模型的装置的结构示意图；
图4是本说明书实施例提供的一种评价改编水平的装置的结构示意图；
图5是用于配置本说明书实施例方法的一种计算机设备的结构示意图。
具体实施方式
为了使本领域技术人员更好地理解本说明书实施例中的技术方案，下面将结合本说明书实施例中的附图，对本说明书实施例中的技术方案进行详细地描述，显然，所描述的实施例仅仅是本说明书的一部分实施例，而不是全部的实施例。基于本说明书中的实施例，本领域普通技术人员所获得的所有其他实施例，都应当属于保护的范围。
以下结合附图，详细说明本说明书各实施例提供的技术方案。
图1是本说明书实施例提供的一种训练改编水平评价模型的方法的流程示意图，包括以下步骤：
S100：获取改编音乐作品样本集合。
本方法的执行主体可以是具有数据处理功能的设备，如服务器、计算机、手机等。
在本说明书实施例中，为了使得训练得到的改编水平评价模型较为准确，一般需要获取大量的改编音乐作品作为样本，组成改编音乐作品样本集合。
需要说明的是，改编音乐作品一般是数字化格式的纯编曲文件，如乐器数字接口(Musical Instrument Digital Interface，MIDI)格式的编曲文件。
S102：针对所述改编音乐作品集合中的每个改编音乐作品样本，分离出该改编音乐作品样本的各音轨，并确定该改编音乐作品样本的各音轨分别对应的演奏乐器信息。
S104：分别对该改编音乐作品样本的各音轨进行乐音原创性分析，得到该改编音乐作品样本的各音轨分别对应的乐音原创性表征值。
S106：针对该改编音乐作品样本的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为该改编音乐作品样本的一个样本特征。
众所周知，在模型训练领域，可以对样本进行特征提取，并针对每个样本，基于该样本的样本特征和该样本的标签，采用有监督学习算法，进行模型训练。模型训练的过程，实际上是机器对样本的样本特征与标签之间的映射关系进行学习的过程。因此，待模型训练完毕后，就可以将待验证的目标对象的特征输入到模型，由模型输出目标对象的标签。
在本说明书实施例中，针对改编音乐作品集合中的每个改编音乐作品样本，为了得到该改编音乐作品样本的样本特征，可以执行步骤S102～S106。
通常，对于一个编曲文件而言，其每个音轨对应的旋律是由一种或多种演奏乐器进行演奏的，各音轨分别对应的演奏乐器不尽相同。在本说明书实施例中，针对改编音乐作品集合中的每个改编音乐作品样本，如果该改编音乐作品样本是MIDI格式的编曲文件，那么，由于MIDI格式的编曲文件的旋律本身就是对各音轨进行分别编排的，并且，该改编音乐作品样本也包含了每个音轨对应的演奏乐器信息，因此，很容易分离出该改编音乐作品样本的各音轨，并获得该改编音乐作品样本的各音轨分别对应的演奏乐器信息。需要说明的是，演奏乐器信息一般是预设的演奏乐器编号，例如，吉他的演奏乐器编号为1，贝斯的演奏乐器编号为5。演奏乐器信息还可以是其他可以唯一标识演奏乐器的字符串。
此外，如果该改编音乐作品样本是除MIDI格式之外的其他数字化格式的编曲文件，那么，可以采用常见的音轨分离技术手段，将该改编音乐作品样本的各音轨分离出来，进而根据各种演奏乐器不同的音色特征，分析出该改编音乐作品样本的每个音轨对应的演奏乐器。需要说明的是，这种情况下，可以预先为每种演奏乐器分配演奏乐器编号，这样，分析出该改编音乐作品样本的每个音轨对应的演奏乐器，也就确定了该改编音乐作品样本的各音轨分别对应的演奏乐器信息。
在本说明书实施例中，乐音原创性分析是指，基于既有的旋律库，采用数字分析手段，将目标旋律与旋律库中的参考旋律的进行比对，从旋律库中确定出与目标旋律最相似的参考旋律，作为重点参考旋律，并得到目标旋律与重点参考旋律的相似度。此是本领域技术人员所熟知的技术手段。
其中，所述相似度一般是介于0～100％之间的值，表征目标旋律与重点参考旋律之间的相似程度。举例来说，所述相似度为20％，表明目标旋律与重点参考旋律有20％的部分相似。显然，所述相似度越高，所述目标旋律的原创程度越低。
在本说明书实施例中，针对改编音乐作品样本集合中的每个改编音乐作品样本，该改编音乐作品样本的每个音轨都对应有一段旋律，于是，对该改编音乐作品样本的每个音轨都进行乐音原创性分析，就可以得到该改编音乐作品样本的每个音轨对应的乐音原创性表征值。其中，针对任一音轨，该音轨对应的乐音原创性表征值可以是该音轨对应的旋律与旋律库中的重点参考旋律的相似度，此时，所述乐音原创性表征值越大，该音轨对应的旋律的原创程度就越低。该音轨对应的乐音原创性表征值也可以是(1-所述相似度)，此时，所述乐音原创性表征值越大，该音轨对应的旋律的原创程度就越大。
此外，该音轨对应的乐音原创性表征值也可以对所述相似度进行其他处理得到的。 总之，音轨对应的乐音原创性表征值可以表征音轨对应的旋律的原创程度即可。
在本说明书实施例中，以所述改编音乐作品集合中的任一改编音乐作品样本为例说明，针对该改编音乐作品样本的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为该改编音乐作品样本的一个样本特征。
例如，针对某个改编音乐作品样本，通过对该改编音乐作品样本进行音轨分离，分离出3个音轨A、B、C，并进一步确定音轨A对应的演奏乐器编号为1(吉他)，音轨B对应的演奏乐器编号为5(贝斯)，音轨C对应的演奏乐器编号为3(架子鼓)。同时，对该改编音乐作品样本的每个音轨进行乐音原创性分析，得到音轨A对应的乐音原创性表征值为70，音轨B对应的乐音原创性表征值为50，音轨C对应的乐音原创性表征值为30。那么，基于步骤S106，可以将(1，70)、(5，50)、(3，30)这三个演奏乐器编号与乐音原创性表征值的组合，作为该改编音乐作品样本的三个样本特征。
S108：以该改编音乐作品样本的各样本特征为模型输入，以已知的，该改编音乐作品样本对应的改编水平评分为模型输出，进行模型训练，以便得到改编水平评价模型。
针对所述改编音乐作品集合中的每个改编音乐作品样本，在得到该改编音乐作品样本的样本特征之后，还需要获取该改编音乐作品样本的标签，即该改编音乐作品样本对应的改编水平评分。通常，该改编音乐作品样本对应的改编水平评分是由人工指定的。具体地，可以是由若干音乐专家，依据主观经验，对该改编音乐作品样本的改编水平进行评价，给出该改编音乐作品样本的改编水平评分。
获取到所述改编音乐作品集合中的每个改编音乐作品样本的样本特征和改编水平评分后，就可以针对每个改编音乐作品样本，以该改编音乐作品样本的各样本特征为模型输入，以该改编音乐作品样本对应的改编水平评分为模型输出，进行模型训练。最终，得到改编水平评价模型。
此外，以所述改编音乐作品样本集合中的任一改编音乐作品样本为例，在步骤S102中，针对该改编音乐作品样本的某个音轨，可能出现以下情况：
情况一，该音轨对应的演奏乐器可能无法被识别；
情况二，该音轨对应的演奏乐器可能是比较少见的演奏乐器。
对于情况一和情况二，可以预设一个指定编号，作为未知演奏乐器编号，用于统一标识无法被识别的演奏乐器和少见的演奏乐器。如此一来，音轨对应的演奏乐器信息有时是预设的演奏乐器编号，有时是指定编号。
例如，所述指定标号可以是99，对于该改编音乐作品样本的音轨A，当无法分析出音轨A对应的演奏乐器时，将音轨A对应的演奏乐器编号确定为99；对于该改编音乐作品样本的音轨B，当分析出的音轨B对应的演奏乐器较为少见，不在预设的若干演奏乐器中时，将音轨B对应的演奏乐器编号也确定为99。对于该改编音乐作品样本的音轨C，分析出音轨C对应的演奏乐器为架子鼓，在预设的若干演奏乐器中，于是将音轨C对应的演奏乐器编号确定为架子鼓对应的编号，如3。
在本说明书实施例中，进一步地，改编音乐作品样本的样本特征可以不仅包括若干演奏乐器信息与乐音原创性表征值的组合，还可以包括下列特征中的至少一种：
1、预先指定的演奏乐器信息的数量。具体地，在进行模型训练之前，针对每个改编音乐作品样本，确定该改编音乐作品样本对应的演奏乐器信息中，预先指定的演奏乐器信息的数量；将确定的数量作为该改编音乐作品样本对应的一个样本特征。
2、改编音乐作品样本对应的协调性表征值。具体地，在进行模型训练之前，针对每个改编音乐作品样本，对该改编音乐作品样本的各音轨进行协调性分析，得到该改编音乐作品样本对应的协调性表征值；将得到的协调性表征值作为该改编音乐作品样本对应的一个样本特征。
其中，对各音轨进行协调性分析，就是分析各音轨分别对应的旋律的节奏是否合拍，改编音乐作品样本的各音轨分别对应的旋律的节奏合拍程度用改编音乐作品样本对应的协调性表征值来表征。
以改编音乐作品样本集合中任一改编音乐作品样本为例说明。假设预先指定的演奏乐器信息为1、5和6。该改编音乐作品样本的音轨A对应的演奏作品编号为1(吉他)，音轨B对应的演奏作品编号为5(贝斯)，音轨C对应的演奏作品编号为99(未知演奏乐器)，可见，该改编音乐作品样本对应的演奏乐器信息中，预先指定的演奏乐器信息(即编号1和编号5)的数量为2。同时，通过乐音原创性分析，音轨A对应的乐音原创性表征值为20，音轨B对应的乐音原创性表征值为60，音轨C对应的乐音原创性表征值为80；通过协调性分析，该改编音乐作品样本对应的协调性表征值为59。那么，该改编音乐作品样本的样本特征为(1，20)、(5，60)、(99，80)、2、59，共5个样本特征。
图2是本说明书实施例提供的一种评价改编水平的方法的流程示意图，包括以下步骤：
S200：获取改编音乐作品。
S202：分离出所述改编音乐作品的各音轨，并确定所述改编音乐作品的各音轨分别对应的演奏乐器信息。
S204：分别对所述改编音乐作品的各音轨进行乐音原创性分析，得到所述改编音乐作品的各音轨分别对应的乐音原创性表征值。
S206：针对所述改编音乐作品的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为所述改编音乐作品的一个特征。
S208：将所述改编音乐作品的各特征输入到改编水平评价模型，以得到所述改编水平评价模型输出的对应于所述改编音乐作品的改编水平评分。
图2所示的评价改编水平的方法，实际上是使用通过图1所示的训练改编水平评价模型的方法得到的改编水平评价模型，对待评价的改编音乐作品进行评价的方法。
对于待评价的改编音乐作品，同样需要得到待评价的改编音乐作品对应的特征，将待评价的改编音乐作品的特征输入到改编水平评价模型，改编水平评价模型会输出该改编音乐作品对应的改编水平评分。一般，改编水平评分越高，说明改编水平越高。
其中，针对所述改编音乐作品的每个音轨，该音轨对应的演奏乐器信息为，对应于已知演奏乐器的编号，或，对应于未知演奏乐器的编号。
在将所述改编音乐作品的各特征输入到改编水平评价模型之前，可以确定所述改编音乐作品对应的演奏乐器信息中，预先指定的演奏乐器信息的数量；将确定的数量作为所述改编音乐作品对应的一个特征。
在将所述改编音乐作品的各样本特征输入到改编水平评价模型之前，可以对所述改编音乐作品的各音轨进行协调性分析，得到所述改编音乐作品对应的协调性表征值；将得到的协调性表征值作为所述改编音乐作品对应的一个特征。
需要说明的是，在图2所示的方法中，得到待评价的改编音乐作品的特征的具体方式，与对图1所示方法的说明中所述的，得到任一改编音乐作品样本的样本特征的方式相同。
通过图1所示的训练改编水平评价模型的方法和图2所示的评价改编水平的方法，利用训练好的改编水平评价模型，可以高效地完成对大量改编音乐作品的改编水平评价工作。并且，以改编音乐作品的每个音轨对应的演奏乐器信息和乐音原创性表征值为特 征进行模型训练，可以使得改编水平评价模型对改编音乐作品的评价水平的评价较为准确。
基于图1所示的训练改编水平评价模型的方法，本说明书实施例还对应提供了一种训练改编水平评价模型的装置，如图3所示，所述装置包括：
获取模块301，获取改编音乐作品样本集合；
音轨分离模块302，针对所述改编音乐作品集合中的每个改编音乐作品样本，分离出该改编音乐作品样本的各音轨，并确定该改编音乐作品样本的各音轨分别对应的演奏乐器信息；
原创性分析模块303，分别对该改编音乐作品样本的各音轨进行乐音原创性分析，得到该改编音乐作品样本的各音轨分别对应的乐音原创性表征值；
确定模块304，针对该改编音乐作品样本的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为该改编音乐作品样本的一个样本特征；
训练模块305，以该改编音乐作品样本的各样本特征为模型输入，以已知的，该改编音乐作品样本对应的改编水平评分为模型输出，进行模型训练，以便得到改编水平评价模型。
针对该改编音乐作品样本的每个音轨，该音轨对应的演奏乐器信息为，预设的演奏乐器编号，或，指定编号。
所述训练模块305，在进行模型训练之前，确定该改编音乐作品样本对应的演奏乐器信息中，预先指定的演奏乐器信息的数量；将确定的数量作为该改编音乐作品样本对应的一个样本特征。
所述训练模块305，在进行模型训练之前，对该改编音乐作品样本的各音轨进行协调性分析，得到该改编音乐作品样本对应的协调性表征值；将得到的协调性表征值作为该改编音乐作品样本对应的一个样本特征。
针对所述改编音乐作品样本集合中的每个改编音乐作品样本，该改编音乐作品样本对应的改编水平评分是由人工指定的。
基于图2所示的评价改编水平的方法，本说明书实施例还对应提供了一种评价改编水平的装置，如图4所示，所述装置包括：
获取模块401，获取改编音乐作品；
音轨分离模块402，分离出所述改编音乐作品的各音轨，并确定所述改编音乐作品的各音轨分别对应的演奏乐器信息；
原创性分析模块403，分别对所述改编音乐作品的各音轨进行乐音原创性分析，得到所述改编音乐作品的各音轨分别对应的乐音原创性表征值；
确定模块404，针对所述改编音乐作品的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为所述改编音乐作品的一个特征；
评价模块405，将所述改编音乐作品的各特征输入到改编水平评价模型，以得到所述改编水平评价模型输出的对应于所述改编音乐作品的改编水平评分；
其中，所述改编水平评价模型是根据图1所示的训练改编水平评价模型的方法得到的。
针对所述改编音乐作品的每个音轨，该音轨对应的演奏乐器信息为，对应于已知演奏乐器的编号，或，对应于未知演奏乐器的编号。
所述评价模块405，在将所述改编音乐作品的各特征输入到改编水平评价模型之前，确定所述改编音乐作品对应的演奏乐器信息中，预先指定的演奏乐器信息的数量；将确定的数量作为所述改编音乐作品对应的一个特征。
所述评价模块405，在将所述改编音乐作品的各样本特征输入到改编水平评价模型之前，对所述改编音乐作品的各音轨进行协调性分析，得到所述改编音乐作品对应的协调性表征值；将得到的协调性表征值作为所述改编音乐作品对应的一个特征。
本说明书实施例还提供一种计算机设备，其至少包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，其中，处理器执行所述程序时实现图1和图2所述方法的功能。
图5示出了本说明书实施例所提供的一种更为具体的计算设备硬件结构示意图，该设备可以包括：处理器1010、存储器1020、输入/输出接口1030、通信接口1040和总线1050。其中处理器1010、存储器1020、输入/输出接口1030和通信接口1040通过总线1050实现彼此之间在设备内部的通信连接。
处理器1010可以采用通用的CPU(Central Processing Unit，中央处理器)、微处理器、应用专用集成电路(Application Specific Integrated Circuit，ASIC)、或者一个或多个集成电路等方式实现，用于执行相关程序，以实现本说明书实施例所提供的技术 方案。
存储器1020可以采用ROM(Read Only Memory，只读存储器)、RAM(Random Access Memory，随机存取存储器)、静态存储设备，动态存储设备等形式实现。存储器1020可以存储操作***和其他应用程序，在通过软件或者固件来实现本说明书实施例所提供的技术方案时，相关的程序代码保存在存储器1020中，并由处理器1010来调用执行。
输入/输出接口1030用于连接输入/输出模块，以实现信息输入及输出。输入输出/模块可以作为组件配置在设备中(图中未示出)，也可以外接于设备以提供相应功能。其中输入设备可以包括键盘、鼠标、触摸屏、麦克风、各类传感器等，输出设备可以包括显示器、扬声器、振动器、指示灯等。
通信接口1040用于连接通信模块(图中未示出)，以实现本设备与其他设备的通信交互。其中通信模块可以通过有线方式(例如USB、网线等)实现通信，也可以通过无线方式(例如移动网络、WIFI、蓝牙等)实现通信。
总线1050包括一通路，在设备的各个组件(例如处理器1010、存储器1020、输入/输出接口1030和通信接口1040)之间传输信息。
需要说明的是，尽管上述设备仅示出了处理器1010、存储器1020、输入/输出接口1030、通信接口1040以及总线1050，但是在具体实施过程中，该设备还可以包括实现正常运行所必需的其他组件。此外，本领域的技术人员可以理解的是，上述设备中也可以仅包含实现本说明书实施例方案所必需的组件，而不必包含图中所示的全部组件。
本说明书实施例还提供一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时实现图1和图2所述方法的功能。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑 可读媒体(transitory media)，如调制的数据信号和载波。
通过以上的实施方式的描述可知，本领域的技术人员可以清楚地了解到本说明书实施例可借助软件加必需的通用硬件平台的方式来实现。基于这样的理解，本说明书实施例的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品可以存储在存储介质中，如ROM/RAM、磁碟、光盘等，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本说明书实施例各个实施例或者实施例的某些部分所述的方法。
上述实施例阐明的***、方法、模块或单元，具体可以由计算机芯片或实体实现，或者由具有某种功能的产品来实现。一种典型的实现设备为计算机，计算机的具体形式可以是个人计算机、膝上型计算机、蜂窝电话、相机电话、智能电话、个人数字助理、媒体播放器、导航设备、电子邮件收发设备、游戏控制台、平板计算机、可穿戴设备或者这些设备中的任意几种设备的组合。
本说明书中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于装置和设备实施例而言，由于其基本相似于方法实施例，所以描述得比较简单，相关之处参见方法实施例的部分说明即可。以上所描述的方法实施例仅仅是示意性的，其中所述作为分离部件说明的模块可以是或者也可以不是物理上分开的，在实施本说明书实施例方案时可以把各模块的功能在同一个或多个软件和/或硬件中实现。也可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性劳动的情况下，即可以理解并实施。
以上所述仅是本说明书实施例的具体实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本说明书实施例原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也应视为本说明书实施例的保护范围。

Claims (20)

  1. 一种训练改编水平评价模型的方法，包括：
    获取改编音乐作品样本集合；
    针对所述改编音乐作品集合中的每个改编音乐作品样本，分离出该改编音乐作品样本的各音轨，并确定该改编音乐作品样本的各音轨分别对应的演奏乐器信息；
    分别对该改编音乐作品样本的各音轨进行乐音原创性分析，得到该改编音乐作品样本的各音轨分别对应的乐音原创性表征值；
    针对该改编音乐作品样本的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为该改编音乐作品样本的一个样本特征；
    以该改编音乐作品样本的各样本特征为模型输入，以已知的，该改编音乐作品样本对应的改编水平评分为模型输出，进行模型训练，以便得到改编水平评价模型。
  2. 如权利要求1所述的方法，针对该改编音乐作品样本的每个音轨，该音轨对应的演奏乐器信息为，预设的演奏乐器编号，或，指定编号。
  3. 如权利要求1所述的方法，在进行模型训练之前，所述方法还包括：
    确定该改编音乐作品样本对应的演奏乐器信息中，预先指定的演奏乐器信息的数量；
    将确定的数量作为该改编音乐作品样本对应的一个样本特征。
  4. 如权利要求1所述的方法，在进行模型训练之前，所述方法还包括：
    对该改编音乐作品样本的各音轨进行协调性分析，得到该改编音乐作品样本对应的协调性表征值；
    将得到的协调性表征值作为该改编音乐作品样本对应的一个样本特征。
  5. 如权利要求1所述的方法，针对所述改编音乐作品样本集合中的每个改编音乐作品样本，该改编音乐作品样本对应的改编水平评分是由人工指定的。
  6. 一种评价改编水平的方法，包括：
    获取改编音乐作品；
    分离出所述改编音乐作品的各音轨，并确定所述改编音乐作品的各音轨分别对应的演奏乐器信息；
    分别对所述改编音乐作品的各音轨进行乐音原创性分析，得到所述改编音乐作品的各音轨分别对应的乐音原创性表征值；
    针对所述改编音乐作品的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为所述改编音乐作品的一个特征；
    将所述改编音乐作品的各特征输入到改编水平评价模型，以得到所述改编水平评价模型 输出的对应于所述改编音乐作品的改编水平评分；
    其中，所述改编水平评价模型是根据权利要求1～4任一项所述的方法得到的。
  7. 如权利要求6所述的方法，针对所述改编音乐作品的每个音轨，该音轨对应的演奏乐器信息为，对应于已知演奏乐器的编号，或，对应于未知演奏乐器的编号。
  8. 如权利要求6所述的方法，在将所述改编音乐作品的各样本特征输入到改编水平评价模型之前，所述方法还包括：
    确定所述改编音乐作品对应的演奏乐器信息中，预先指定的演奏乐器信息的数量；
    将确定的数量作为所述改编音乐作品对应的一个特征。
  9. 如权利要求6所述的方法，在将所述改编音乐作品的各样本特征输入到改编水平评价模型之前，所述方法还包括：
    对所述改编音乐作品的各音轨进行协调性分析，得到所述改编音乐作品对应的协调性表征值；
    将得到的协调性表征值作为所述改编音乐作品对应的一个特征。
  10. 一种训练改编水平评价模型的装置，包括：
    获取模块，获取改编音乐作品样本集合；
    音轨分离模块，针对所述改编音乐作品集合中的每个改编音乐作品样本，分离出该改编音乐作品样本的各音轨，并确定该改编音乐作品样本的各音轨分别对应的演奏乐器信息；
    原创性分析模块，分别对该改编音乐作品样本的各音轨进行乐音原创性分析，得到该改编音乐作品样本的各音轨分别对应的乐音原创性表征值；
    确定模块，针对该改编音乐作品样本的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为该改编音乐作品样本的一个样本特征；
    训练模块，以该改编音乐作品样本的各样本特征为模型输入，以已知的，该改编音乐作品样本对应的改编水平评分为模型输出，进行模型训练，以便得到改编水平评价模型。
  11. 如权利要求10所述的装置，针对该改编音乐作品样本的每个音轨，该音轨对应的演奏乐器信息为，预设的演奏乐器编号，或，指定编号。
  12. 如权利要求10所述的装置，所述训练模块，在进行模型训练之前，确定该改编音乐作品样本对应的演奏乐器信息中，预先指定的演奏乐器信息的数量；将确定的数量作为该改编音乐作品样本对应的一个样本特征。
  13. 如权利要求10所述的装置，所述训练模块，在进行模型训练之前，对该改编音乐作品样本的各音轨进行协调性分析，得到该改编音乐作品样本对应的协调性表征值；将得到的协调性表征值作为该改编音乐作品样本对应的一个样本特征。
  14. 如权利要求10所述的装置，针对所述改编音乐作品样本集合中的每个改编音乐作品样本，该改编音乐作品样本对应的改编水平评分是由人工指定的。
  15. 一种评价改编水平的装置，包括：
    获取模块，获取改编音乐作品；
    音轨分离模块，分离出所述改编音乐作品的各音轨，并确定所述改编音乐作品的各音轨分别对应的演奏乐器信息；
    原创性分析模块，分别对所述改编音乐作品的各音轨进行乐音原创性分析，得到所述改编音乐作品的各音轨分别对应的乐音原创性表征值；
    确定模块，针对所述改编音乐作品的每个音轨，将该音轨对应的演奏乐器信息与乐音原创性表征值的组合，作为所述改编音乐作品的一个特征；
    评价模块，将所述改编音乐作品的各样本特征输入到改编水平评价模型，以得到所述改编水平评价模型输出的对应于所述改编音乐作品的改编水平评分；
    其中，所述改编水平评价模型是根据权利要求1～4任一项所述的方法得到的。
  16. 如权利要求15所述的装置，针对所述改编音乐作品的每个音轨，该音轨对应的演奏乐器信息为，对应于已知演奏乐器的编号，或，对应于未知演奏乐器的编号。
  17. 如权利要求15所述的装置，所述评价模块，在将所述改编音乐作品的各样本特征输入到改编水平评价模型之前，确定所述改编音乐作品对应的演奏乐器信息中，预先指定的演奏乐器信息的数量；将确定的数量作为所述改编音乐作品对应的一个特征。
  18. 如权利要求15所述的装置，所述评价模块，在将所述改编音乐作品的各样本特征输入到改编水平评价模型之前，对所述改编音乐作品的各音轨进行协调性分析，得到所述改编音乐作品对应的协调性表征值；将得到的协调性表征值作为所述改编音乐作品对应的一个特征。
  19. 一种计算机设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，其中，所述处理器执行所述程序时实现如权利要求1～5任一项所述的方法。
  20. 一种计算机设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，其中，所述处理器执行所述程序时实现如权利要求6～9任一项所述的方法。
PCT/CN2019/083092 2018-07-18 2019-04-17 一种训练改编水平评价模型、评价改编水平的方法及装置 WO2020015411A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
EP19838690.6A EP3779814A4 (en) 2018-07-18 2019-04-17 METHOD AND DEVICE FOR TRAINING AN ADAPTATION LEVEL ASSESSMENT MODEL, AND ADAPTATION LEVEL ASSESSMENT METHOD AND DEVICE
SG11202010627QA SG11202010627QA (en) 2018-07-18 2019-04-17 Method and apparatus for training adaptation quality evaluation model, and method and apparatus for evaluating adaptation quality
US17/083,740 US11074897B2 (en) 2018-07-18 2020-10-29 Method and apparatus for training adaptation quality evaluation model, and method and apparatus for evaluating adaptation quality
US17/382,860 US11367424B2 (en) 2018-07-18 2021-07-22 Method and apparatus for training adaptation quality evaluation model, and method and apparatus for evaluating adaptation quality

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810793081.3 2018-07-18
CN201810793081.3A CN109190879B (zh) 2018-07-18 2018-07-18 一种训练改编水平评价模型、评价改编水平的方法及装置

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/083,740 Continuation US11074897B2 (en) 2018-07-18 2020-10-29 Method and apparatus for training adaptation quality evaluation model, and method and apparatus for evaluating adaptation quality

Publications (1)

Publication Number Publication Date
WO2020015411A1 true WO2020015411A1 (zh) 2020-01-23

Family

ID=64936255

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/083092 WO2020015411A1 (zh) 2018-07-18 2019-04-17 一种训练改编水平评价模型、评价改编水平的方法及装置

Country Status (6)

Country Link
US (2) US11074897B2 (zh)
EP (1) EP3779814A4 (zh)
CN (1) CN109190879B (zh)
SG (1) SG11202010627QA (zh)
TW (1) TWI703515B (zh)
WO (1) WO2020015411A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113780811A (zh) * 2021-09-10 2021-12-10 平安科技（深圳）有限公司 乐器演奏评估方法、装置、设备及存储介质
CN115132183A (zh) * 2022-05-25 2022-09-30 腾讯科技（深圳）有限公司 音频识别模型的训练方法、装置、设备、介质及程序产品

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109190879B (zh) * 2018-07-18 2020-08-11 阿里巴巴集团控股有限公司 一种训练改编水平评价模型、评价改编水平的方法及装置
GB2595222A (en) * 2020-05-18 2021-11-24 Waves Audio Ltd Digital audio workstation with audio processing recommendations

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN201294089Y (zh) * 2008-11-17 2009-08-19 音乐传奇有限公司 互动音乐演奏设备
US20110011243A1 (en) * 2009-07-20 2011-01-20 Apple Inc. Collectively adjusting tracks using a digital audio workstation
CN105590633A (zh) * 2015-11-16 2016-05-18 福建省百利亨信息科技有限公司 一种用于歌曲评分的曲谱生成方法和设备
CN106652655A (zh) * 2015-10-29 2017-05-10 施政 一种音轨替换的乐器
CN109190879A (zh) * 2018-07-18 2019-01-11 阿里巴巴集团控股有限公司 一种训练改编水平评价模型、评价改编水平的方法及装置

Family Cites Families (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TW316961B (en) * 1995-04-18 1997-10-01 Texas Instruments Inc Method and system for scoring a musical performance
CN2271006Y (zh) * 1995-09-19 1997-12-17 史耀华 音乐助教或娱乐装置
US7723602B2 (en) * 2003-08-20 2010-05-25 David Joseph Beckford System, computer program and method for quantifying and analyzing musical intellectual property
US8707459B2 (en) * 2007-01-19 2014-04-22 Digimarc Corporation Determination of originality of content
US9654447B2 (en) * 2006-08-29 2017-05-16 Digimarc Corporation Customized handling of copied content based on owner-specified similarity thresholds
US8924270B2 (en) * 2007-05-03 2014-12-30 Google Inc. Monetization of digital content contributions
JP4953478B2 (ja) * 2007-07-31 2012-06-13 独立行政法人産業技術総合研究所 楽曲推薦システム、楽曲推薦方法及び楽曲推薦用コンピュータプログラム
TWI410811B (zh) * 2009-09-08 2013-10-01 Univ Nat Cheng Kung 音樂推薦方法與其程式產品
JP6248415B2 (ja) * 2013-05-23 2017-12-20 ヤマハ株式会社 楽曲評価装置
WO2016102737A1 (en) * 2014-12-22 2016-06-30 Nokia Technologies Oy Tagging audio data
CN104680335A (zh) * 2015-03-24 2015-06-03 北京闪信鼎中技术有限公司 一种新能源汽车的测评方法
US20170134162A1 (en) * 2015-11-10 2017-05-11 Shannon Code System and process for verifying digital media content authenticity
US11544670B2 (en) * 2016-08-07 2023-01-03 Verifi Media, Inc. Distributed data store for managing media
CN106682695A (zh) * 2016-12-28 2017-05-17 中国农业大学 一种基于支持向量机的县域耕地自然质量评价方法
CN107845408B (zh) * 2017-10-25 2020-10-27 医渡云（北京）技术有限公司 数据评价方法及装置、存储介质及电子设备

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN201294089Y (zh) * 2008-11-17 2009-08-19 音乐传奇有限公司 互动音乐演奏设备
US20110011243A1 (en) * 2009-07-20 2011-01-20 Apple Inc. Collectively adjusting tracks using a digital audio workstation
CN106652655A (zh) * 2015-10-29 2017-05-10 施政 一种音轨替换的乐器
CN105590633A (zh) * 2015-11-16 2016-05-18 福建省百利亨信息科技有限公司 一种用于歌曲评分的曲谱生成方法和设备
CN109190879A (zh) * 2018-07-18 2019-01-11 阿里巴巴集团控股有限公司 一种训练改编水平评价模型、评价改编水平的方法及装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3779814A4 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113780811A (zh) * 2021-09-10 2021-12-10 平安科技（深圳）有限公司 乐器演奏评估方法、装置、设备及存储介质
CN113780811B (zh) * 2021-09-10 2023-12-26 平安科技（深圳）有限公司 乐器演奏评估方法、装置、设备及存储介质
CN115132183A (zh) * 2022-05-25 2022-09-30 腾讯科技（深圳）有限公司 音频识别模型的训练方法、装置、设备、介质及程序产品
CN115132183B (zh) * 2022-05-25 2024-04-12 腾讯科技（深圳）有限公司 音频识别模型的训练方法、装置、设备、介质及程序产品

Also Published As

Publication number Publication date
US11367424B2 (en) 2022-06-21
TW202006613A (zh) 2020-02-01
US11074897B2 (en) 2021-07-27
EP3779814A4 (en) 2021-04-28
CN109190879A (zh) 2019-01-11
TWI703515B (zh) 2020-09-01
SG11202010627QA (en) 2020-11-27
US20210043173A1 (en) 2021-02-11
US20210350775A1 (en) 2021-11-11
EP3779814A1 (en) 2021-02-17
CN109190879B (zh) 2020-08-11

Similar Documents

Publication Publication Date Title
WO2020015411A1 (zh) 一种训练改编水平评价模型、评价改编水平的方法及装置
TWI719557B (zh) 基於梯度提升決策樹的模型訓練方法及裝置
US9576050B1 (en) Generating a playlist based on input acoustic information
JP2009139769A (ja) 信号処理装置、信号処理方法及びプログラム
CN110010159B (zh) 声音相似度确定方法及装置
US11271993B2 (en) Streaming music categorization using rhythm, texture and pitch
JP2019219638A (ja) 音楽合成方法及びシステム、端末並びにコンピュータ可読記憶媒体
US20210049509A1 (en) Blockchain-based method and apparatus for rewarding work performer
JP6151121B2 (ja) コード進行推定検出装置及びコード進行推定検出プログラム
JP2021101252A (ja) 情報処理方法、情報処理装置およびプログラム
JP2019144485A (ja) コード特定方法およびプログラム
KR101813704B1 (ko) 사용자 음색 분석 장치 및 음색 분석 방법
CN106775567B (zh) 一种音效匹配方法及***
WO2024001548A1 (zh) 歌单生成方法、装置、电子设备及存储介质
JP2006155157A5 (zh)
JP2017027070A (ja) 評価装置及びプログラム
KR101467852B1 (ko) 코드 이미지가 인쇄된 인쇄물을 이용하는 전자펜의 악기 연주 음향재생 제어방법 및 이를 수행하는 전자펜
JP7428182B2 (ja) 情報処理装置および方法、並びにプログラム
CN116189636B (zh) 基于电子乐器的伴奏生成方法、装置、设备及存储介质
US11609948B2 (en) Music streaming, playlist creation and streaming architecture
JP6726583B2 (ja) 情報処理装置、情報処理システム、情報処理方法、及びプログラム
JP2022096204A (ja) 譜面生成装置及びプログラム
TWI512500B (zh) 調整多媒體裝置之數位訊號處理設定之方法及系統，及其電腦程式產品
CN115019753A (zh) 音频处理方法及装置、电子设备、计算机可读存储介质
TW202420143A (zh) 媒體檔案選擇方法及服務系統與電腦程式產品

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19838690

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 19838690.6

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2019838690

Country of ref document: EP

Effective date: 20201030

NENP Non-entry into the national phase

Ref country code: DE