Music generation method and device based on intelligent voice analysis

Abstract

The invention relates to the technical field of data processing, and discloses a music generation method and device based on intelligent voice analysis, and the method comprises the steps: obtaining a target voice input by a user, carrying out the voice analysis of the target voice, obtaining the voice content and voice emotion corresponding to the target voice, and extracting the voice content according to a preset content extraction rule, and extracting at least one keyword in the voice content, analyzing the voice emotion to obtain at least one emotion label corresponding to the voice emotion, and generating a target song corresponding to the target voice according to each keyword and each emotion label. It can be seen that the corresponding songs can be generated according to the content and emotion taught by the user, the song generation difficulty is reduced, the song generation interestingness is improved, the user can create unique music in the taught process, the interest of the user in the music is improved, and the use experience of the user is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a music generation method and device based on intelligent speech analysis. Background Art

[0002] As an art form and cultural activity that can convey and express emotions, music can have a positive impact on children. Learning music can enrich children's emotional experience, subtly infect and edify their emotional world, and gradually establish a love for humanity, nature, and all beautiful things. Therefore, more and more parents are beginning to let their children get in touch with and learn music.

[0003] However, the process of learning and mastering music is relatively boring, especially the process of song creation, which requires not only more professional music theory knowledge, but also creative ideas and inspiration. For children, the creative process is difficult and boring. Therefore, it is particularly important to propose a technical solution that can reduce the difficulty of song generation and increase the fun of song generation. Summary of the invention

[0004] The present invention provides a music generation method and device based on intelligent speech analysis, which can help reduce the difficulty of song generation and improve the interest of song generation.

[0005] In order to solve the above technical problems, the first aspect of the present invention discloses a music generation method based on intelligent speech analysis, the method comprising:

[0006] Acquire a target voice input by a user, and perform voice analysis on the target voice to obtain voice information corresponding to the target voice, wherein the voice information includes voice content and voice emotion;

[0007] Extracting at least one keyword from the speech content according to a preset content extraction rule;

[0008] Analyze the speech emotion to obtain at least one emotion label corresponding to the speech emotion;

[0009] According to each of the keywords and each of the emotion tags, a target song corresponding to the target speech is generated.

[0010] As an optional embodiment, in the first aspect of the present invention, the method further comprises:

[0011] Acquire a network song collection, wherein the network song collection includes a network lyrics collection and a network melody collection;

[0012] Classifying the network lyrics collection according to a preset lyrics classification rule to obtain at least one lyrics sub-database corresponding to the network lyrics collection;

[0013] Performing melody analysis on each network melody in the network melody set to determine at least one melody emotion label corresponding to each network melody;

[0014] For each of the network melodies, melody classification is performed on the network melody according to each of the melody emotion tags corresponding to the network melody, so as to obtain at least one melody sub-database corresponding to the network melody set;

[0015] A song database is generated based on each of the lyrics sub-databases and each of the melody sub-databases.

[0016] As an optional implementation, in the first aspect of the present invention, the target song includes target lyrics and target melody;

[0017] The step of generating a target song corresponding to the target speech according to each of the keywords and each of the emotion tags comprises:

[0018] For each of the keywords, determining the lyrics classification information corresponding to the keyword according to the lyrics classification rule, and determining at least one of the lyrics sub-databases corresponding to the keyword according to the lyrics classification information;

[0019] For each of the keywords, in each of the lyrics sub-databases corresponding to the keyword, screening at least one lyrics segment corresponding to the keyword;

[0020] For each of the emotion tags, determining at least one of the melody emotion tags corresponding to the emotion tag, and according to each of the melody emotion tags, determining at least one of the melody sub-databases corresponding to the emotion tag;

[0021] For each of the emotion tags, in each of the melody sub-databases corresponding to the emotion tag, screening at least one melody segment corresponding to the emotion tag;

[0022] A target song corresponding to the target voice is generated according to each of the lyrics segments and each of the melody segments.

[0023] As an optional implementation manner, in the first aspect of the present invention, generating a target song corresponding to the target speech according to each of the lyrics segments and each of the melody segments includes:

[0024] Combining each of the lyrics fragments to generate at least one combined lyric, and combining each of the melody fragments to generate at least one combined melody;

[0025] For each of the combined lyrics, the combined lyrics are matched with each of the combined melodies to obtain a combined song of the combined lyrics and each of the combined melodies, and a song score of each of the combined songs is determined;

[0026] The song creation habit information corresponding to the user is obtained, and according to the song creation habit information and the song score of each of the combination songs, a target song corresponding to the target voice is determined in each of the combination songs.

[0027] As an optional embodiment, in the first aspect of the present invention, the method further comprises:

[0028] Determining story information corresponding to the voice content, the story information including object information of at least one story object;

[0029] Acquire at least one target story object selected by the user in each of the story objects, and determine a first song optimization parameter according to the object information of each of the target story objects;

[0030] Determine song playing scene information corresponding to the target song, and determine a second song optimization parameter according to the song playing scene information, wherein the song playing scene information includes a scene type and a scene atmosphere;

[0031] The target song is optimized according to the first song optimization parameter and the second song optimization parameter to obtain an optimized target song.

[0032] As an optional embodiment, in the first aspect of the present invention, the method further comprises:

[0033] Performing originality analysis on the target song to obtain an originality score of the target song, and determining whether the originality score is less than a preset originality score threshold to obtain a determination result;

[0034] When the judgment result indicates that the originality score is less than the originality score threshold, determining the infringement risk of the target song, the infringement risk including the lyrics infringement risk of each of the lyrics segments and / or the melody infringement risk of each of the melody segments;

[0035] According to the infringement risk, a song modification plan for the target song is determined, and the song modification plan includes a lyrics modification plan and / or a melody modification plan.

[0036] As an optional implementation, in the first aspect of the present invention, determining a song modification plan for the target song according to the infringement risk includes:

[0037] For each of the lyrics segments, when the lyrics infringement risk of the lyrics segment is greater than a preset lyrics infringement risk threshold, the lyrics segment is determined to be a risky lyrics segment;

[0038] For each of the risky lyrics fragments, determine the lyrics copyright information of the risky lyrics fragment, and judge whether the lyrics copyright information of the risky lyrics fragment meets the lyrics copyright acquisition conditions of the user; when the lyrics copyright information of the risky lyrics fragment meets the lyrics copyright acquisition conditions of the user, determine a lyrics copyright acquisition plan for the risky lyrics fragment; when the lyrics copyright information of the risky lyrics fragment does not meet the lyrics copyright acquisition conditions of the user, determine at least one replacement lyrics fragment corresponding to the risky lyrics fragment in the song database, and determine a lyrics replacement plan for the risky lyrics fragment based on each of the replacement lyrics fragments;

[0039] Determine a lyrics modification plan based on a lyrics copyright acquisition plan or a lyrics replacement plan for each of the risky lyrics segments; and / or,

[0040] For each of the melody segments, when the melody infringement risk of the melody segment is greater than a preset melody infringement risk threshold, the melody segment is determined to be a risky melody segment;

[0041] For each of the risky melody segments, determine the melody copyright information of the risky melody segment, and judge whether the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user; when the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user, determine a melody copyright acquisition scheme for the risky melody segment; when the melody copyright information of the risky melody segment does not meet the melody copyright acquisition condition of the user, determine at least one replacement melody segment corresponding to the risky melody segment in the song database, and determine a melody replacement scheme for the risky melody segment according to each of the replacement melody segments;

[0042] Determine a melody modification plan according to a melody copyright acquisition plan or a melody replacement plan of each of the risky melody fragments;

[0043] A song modification scheme for the target song is determined according to the lyrics modification scheme and / or the melody modification scheme.

[0044] A second aspect of the present invention discloses a music generation device based on intelligent speech analysis, the device comprising:

[0045] An acquisition module is used to acquire a target voice input by a user, and perform voice analysis on the target voice to obtain voice information corresponding to the target voice, wherein the voice information includes voice content and voice emotion;

[0046] An extraction module, used to extract at least one keyword from the speech content according to a preset content extraction rule;

[0047] An analysis module, used to analyze the speech emotion and obtain at least one emotion label corresponding to the speech emotion;

[0048] A generation module is used to generate a target song corresponding to the target speech according to each of the keywords and each of the emotion tags.

[0049] As an optional implementation, in the second aspect of the present invention, the acquisition module is further used to acquire a network song collection, wherein the network song collection includes a network lyrics collection and a network melody collection;

[0050] The device also includes:

[0051] A classification module, used to classify the network lyrics collection according to a preset lyrics classification rule, and obtain at least one lyrics sub-database corresponding to the network lyrics collection;

[0052] The analysis module is further used to perform melody analysis on each network melody in the network melody set to determine at least one melody emotion label corresponding to each network melody;

[0053] The classification module is further used to classify each of the network melodies according to each of the melody emotion tags corresponding to the network melody, so as to obtain at least one melody sub-database corresponding to the network melody set;

[0054] The generating module is further used to generate a song database according to each of the lyrics sub-databases and each of the melody sub-databases.

[0055] As an optional implementation, in the second aspect of the present invention, the target song includes target lyrics and target melody;

[0056] The generation module generates a target song corresponding to the target speech according to each of the keywords and each of the emotion tags, specifically including:

[0057] For each of the keywords, determining the lyrics classification information corresponding to the keyword according to the lyrics classification rule, and determining at least one of the lyrics sub-databases corresponding to the keyword according to the lyrics classification information;

[0058] For each of the keywords, in each of the lyrics sub-databases corresponding to the keyword, screening at least one lyrics segment corresponding to the keyword;

[0059] For each of the emotion tags, determining at least one of the melody emotion tags corresponding to the emotion tag, and according to each of the melody emotion tags, determining at least one of the melody sub-databases corresponding to the emotion tag;

[0060] For each of the emotion tags, in each of the melody sub-databases corresponding to the emotion tag, screening at least one melody segment corresponding to the emotion tag;

[0061] A target song corresponding to the target voice is generated according to each of the lyrics segments and each of the melody segments.

[0062] As an optional implementation, in the second aspect of the present invention, the generation module generates a target song corresponding to the target voice according to each of the lyrics segments and each of the melody segments, specifically including:

[0063] Combining each of the lyrics fragments to generate at least one combined lyric, and combining each of the melody fragments to generate at least one combined melody;

[0064] For each of the combined lyrics, the combined lyrics are matched with each of the combined melodies to obtain a combined song of the combined lyrics and each of the combined melodies, and a song score of each of the combined songs is determined;

[0065] The song creation habit information corresponding to the user is obtained, and according to the song creation habit information and the song score of each of the combination songs, a target song corresponding to the target voice is determined in each of the combination songs.

[0066] As an optional implementation, in the second aspect of the present invention, the device further includes:

[0067] A first determining module, configured to determine story information corresponding to the voice content, wherein the story information includes object information of at least one story object;

[0068] The acquisition module is further used to acquire at least one target story object selected by the user in each of the story objects, and determine the first song optimization parameter according to the object information of each of the target story objects;

[0069] The first determination module is further used to determine song playback scene information corresponding to the target song, and determine the second song optimization parameter according to the song playback scene information, wherein the song playback scene information includes a scene type and a scene atmosphere;

[0070] The optimization module is used to optimize the target song according to the first song optimization parameter and the second song optimization parameter to obtain an optimized target song.

[0071] As an optional implementation, in the second aspect of the present invention, the analysis module is further used to perform originality analysis on the target song to obtain an originality score of the target song, and to determine whether the originality score is less than a preset originality score threshold to obtain a determination result;

[0072] The device also includes:

[0073] A second determination module is configured to determine the infringement risk of the target song when the judgment result indicates that the originality score is less than the originality score threshold, wherein the infringement risk includes the lyrics infringement risk of each of the lyrics segments and / or the melody infringement risk of each of the melody segments;

[0074] The second determination module is further used to determine a song modification plan for the target song according to the infringement risk, and the song modification plan includes a lyrics modification plan and / or a melody modification plan.

[0075] As an optional implementation, in the second aspect of the present invention, the second determination module determines the song modification scheme for the target song according to the infringement risk, specifically including:

[0076] For each of the lyrics segments, when the lyrics infringement risk of the lyrics segment is greater than a preset lyrics infringement risk threshold, the lyrics segment is determined to be a risky lyrics segment;

[0077] For each of the risky lyrics fragments, determine the lyrics copyright information of the risky lyrics fragment, and judge whether the lyrics copyright information of the risky lyrics fragment meets the lyrics copyright acquisition conditions of the user; when the lyrics copyright information of the risky lyrics fragment meets the lyrics copyright acquisition conditions of the user, determine a lyrics copyright acquisition plan for the risky lyrics fragment; when the lyrics copyright information of the risky lyrics fragment does not meet the lyrics copyright acquisition conditions of the user, determine at least one replacement lyrics fragment corresponding to the risky lyrics fragment in the song database, and determine a lyrics replacement plan for the risky lyrics fragment based on each of the replacement lyrics fragments;

[0078] Determine a lyrics modification plan based on a lyrics copyright acquisition plan or a lyrics replacement plan for each of the risky lyrics segments; and / or,

[0079] For each of the melody segments, when the melody infringement risk of the melody segment is greater than a preset melody infringement risk threshold, the melody segment is determined to be a risky melody segment;

[0080] For each of the risky melody segments, determine the melody copyright information of the risky melody segment, and judge whether the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user; when the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user, determine a melody copyright acquisition scheme for the risky melody segment; when the melody copyright information of the risky melody segment does not meet the melody copyright acquisition condition of the user, determine at least one replacement melody segment corresponding to the risky melody segment in the song database, and determine a melody replacement scheme for the risky melody segment according to each of the replacement melody segments;

[0081] Determine a melody modification plan according to a melody copyright acquisition plan or a melody replacement plan of each of the risky melody fragments;

[0082] A song modification scheme for the target song is determined according to the lyrics modification scheme and / or the melody modification scheme.

[0083] The third aspect of the present invention discloses another music generation device based on intelligent speech analysis, the device comprising:

[0084] A memory storing executable program code;

[0085] a processor coupled to the memory;

[0086] The processor calls the executable program code stored in the memory to execute the music generation method based on intelligent speech analysis disclosed in the first aspect of the present invention.

[0087] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the music generation method based on intelligent speech analysis disclosed in the first aspect of the present invention.

[0088] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0089] In the embodiment of the present invention, the target voice input by the user is obtained, and the target voice is subjected to voice analysis to obtain the voice content and voice emotion corresponding to the target voice. According to the preset content extraction rules, at least one keyword in the voice content is extracted, and the voice emotion is analyzed to obtain at least one emotion tag corresponding to the voice emotion. According to each keyword and each emotion tag, a target song corresponding to the target voice is generated. It can be seen that the implementation of the present invention can generate corresponding songs according to the content and emotions narrated by the user, reduce the difficulty of song generation, improve the fun of song generation, enable users to create unique music in the narration process, increase users' interest in music, and enhance users' experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0091] Figure 1 It is a flowchart of a music generation method based on intelligent speech analysis disclosed in an embodiment of the present invention;

[0092] Figure 2 It is a flowchart of another music generation method based on intelligent speech analysis disclosed in an embodiment of the present invention;

[0093] Figure 3 It is a structural schematic diagram of a music generation device based on intelligent speech analysis disclosed in an embodiment of the present invention;

[0094] Figure 4 is a schematic structural diagram of another music generation device based on intelligent speech analysis disclosed in an embodiment of the present invention;

[0095] Figure 5 It is a structural schematic diagram of another music generation device based on intelligent speech analysis disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0096] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0097] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0098] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0099] The present invention discloses a music generation method and device based on intelligent speech analysis, which can generate corresponding songs according to the content and emotions of the user's narration, reduce the difficulty of song generation, improve the interest of song generation, enable the user to create unique music during the narration process, increase the user's interest in music, and enhance the user's experience. The following are detailed descriptions.

[0100] Embodiment 1

[0101] See also Figure 1 , Figure 1 : is a flowchart of a music generation method based on intelligent speech analysis disclosed in an embodiment of the present invention. Figure 1 The described music generation method based on intelligent speech analysis can be applied to a music generation device based on intelligent speech analysis, wherein the music generation device based on intelligent speech analysis may include a music generation device for analyzing speech and generating music, and the music generation device may include a display screen, a button, and an intelligent server or intelligent platform for analyzing speech and generating music, and the intelligent server includes a local server or a cloud server, which is not limited in the embodiments of the present invention. Figure 1 As shown, the music generation method based on intelligent speech analysis may include the following operations:

[0102] 101. Obtain a target voice input by a user, and perform voice analysis on the target voice to obtain voice information corresponding to the target voice, where the voice information includes voice content and voice emotion.

[0103] In an embodiment of the present invention, optionally, a target voice input by a user can be obtained through a music generating device, and the music generating device can include a story composing machine. Specifically, the target voice can be obtained through a voice acquisition device built into the story composing machine, or through a voice acquisition device external to the story composing machine. The voice acquisition device can include a microphone, and the story composing machine can include a touch screen and / or physical buttons. The user can operate on the touch screen or physical buttons. After the user inputs a confirmation instruction to start sound collection or voice recording, the target voice input by the user is collected. The user can include a child, and the target voice can include a story told by the child, but the present invention does not limit this.

[0104] In an embodiment of the present invention, optionally, performing speech analysis on the target speech may include speech content analysis and / or speech emotion analysis. The speech information corresponding to the target speech may include speech content and speech emotion. The speech content may include content narrated by the user, environmental sound content, etc. The content narrated by the user may include story content. The story content may include virtual story content and / or actual story content, such as the user's related experiences. The speech emotion may include the emotion of the user when the user inputs the target speech, and may also include the emotion of at least one target object in the speech content, such as the emotion of a character in the story. The speech emotion may include one or more of sadness, anger, happiness, etc., which are not limited by the present invention.

[0105] 102. Extract at least one keyword from the voice content according to a preset content extraction rule.

[0106] In an embodiment of the present invention, optionally, the preset content extraction rules may be content extraction rules pre-set in the system of the story composer, such as one or more of object name extraction rules, verb extraction rules, adjective extraction rules, etc., or may be automatically generated based on the music generation focus input by the user. For example, after performing speech analysis on the target speech and obtaining the speech content, the key content in the speech content, such as at least one object in the speech content, is displayed to the user, and the content extraction rules are automatically generated based on the object selected by the user. The present invention does not make any limitation.

[0107] In an embodiment of the present invention, optionally, at least one keyword extracted from the voice content can be classified and extracted according to the object, story time period, scene, etc. in the voice content. For example, classification and extraction can be performed according to the object, that is, keywords corresponding to each object in the voice content are extracted, such as one or more of the object's emotions, events, results, etc., and the present invention is not limited thereto.

[0108] 103. Analyze the speech emotion to obtain at least one emotion label corresponding to the speech emotion.

[0109] In an embodiment of the present invention, optionally, at least one emotion label corresponding to the voice emotion may include one or more of admiration, worship, appreciation, entertainment, anxiety, awe, embarrassment, boredom, calmness, confusion, desire, disgust, pain, fascination, jealousy, excitement, fear, hatred, fun, happiness, nostalgia, romance, sadness, satisfaction, sexual desire, sympathy and contentment.

[0110] 104. Generate a target song corresponding to the target speech according to each keyword and each emotion tag.

[0111] In the embodiment of the present invention, optionally, target lyrics and target melody can be generated according to each keyword and each emotion tag, and a target song corresponding to the target voice can be generated according to the target lyrics and target melody.

[0112] It can be seen that implementation Figure 1 The described music generation method based on intelligent speech analysis can obtain the target speech input by the user, and perform speech analysis on the target speech to obtain the speech content and speech emotion corresponding to the target speech, extract at least one keyword in the speech content according to the preset content extraction rules, analyze the speech emotion, and obtain at least one emotion tag corresponding to the speech emotion, generate a target song corresponding to the target speech according to each keyword and each emotion tag, and can generate corresponding songs according to the content and emotions narrated by the user, reduce the difficulty of song generation, and improve the fun of song generation, so that the user can create unique music in the narration process, increase the user's interest in music, and enhance the user experience.

[0113] In an optional embodiment, the music generation method based on intelligent speech analysis may further include the following operations:

[0114] Obtaining a collection of online songs, the online song collection including an online lyrics collection and an online melody collection;

[0115] Classifying the network lyrics collection according to the preset lyrics classification rules to obtain at least one lyrics sub-database corresponding to the network lyrics collection;

[0116] Performing melody analysis on each network melody in the network melody set to determine at least one melody emotion label corresponding to each network melody;

[0117] For each network melody, classify the network melody according to each melody emotion label corresponding to the network melody, and obtain at least one melody sub-database corresponding to the network melody set;

[0118] A song database is generated based on each lyrics sub-database and each melody sub-database.

[0119] In this optional embodiment, optionally, the online song collection may include an online lyrics collection and an online melody collection, wherein the online lyrics collection may include at least one lyrics segment, which may be a complete lyric or a segment of lyrics, and there is no limit on the length of the lyrics segment; the online melody collection may include at least one melody segment, which may be a complete melody or a segment of melody, and there is no limit on the length of the melody segment.

[0120] In this optional embodiment, optionally, the preset lyrics classification rules can be automatically generated based on existing lyrics classification rules on the Internet, or can be pre-set in the system by the developer based on information such as user usage scenarios, and the lyrics of the online lyrics collection are classified to obtain at least one lyrics sub-database corresponding to the online lyrics collection, wherein the lyrics sub-database can include one or more of a narrative lyrics sub-database, a lyrical lyrics sub-database, a descriptive lyrics sub-database, an argumentative lyrics sub-database, and a duet lyrics sub-database, etc., which is not limited in this embodiment.

[0121] In this optional embodiment, optionally, the melody may include an upward melody, an undulating melody and a downward melody, wherein the melody emotion tag of the upward melody may include one or more of advancing, rising, masculine, fierce, etc., the melody emotion tag of the undulating melody may include one or more of lyrical euphemism, lingering, jumping in, exciting, progressive and smooth, etc., the melody emotion tag of the downward melody may include one or more of gloomy, sad, etc., the melody sub-database may include a sub-database corresponding to each melody emotion tag, a network melody may correspond to at least one melody emotion tag, that is, a network melody may appear in at least one melody sub-database, and a song database is generated according to each lyrics sub-database and each melody sub-database. Further optionally, after generating a target song corresponding to the target voice according to each keyword and each emotion tag, the lyrics and melody corresponding to the target song can be updated to the song database, which is not limited in this embodiment.

[0122] It can be seen that implementing this optional embodiment can obtain a collection of online songs, which includes a collection of online lyrics and a collection of online melody. According to preset lyrics classification rules, the lyrics of the online lyrics collection are classified to obtain at least one lyrics sub-database corresponding to the online lyrics collection. Melody analysis is performed on each online melody in the online melody collection to determine at least one melody emotion tag corresponding to each online melody. For each online melody, the melody is classified according to each melody emotion tag corresponding to the online melody to obtain at least one melody sub-database corresponding to the online melody collection. A song database is generated according to each lyrics sub-database and each melody sub-database. The corresponding lyrics sub-database and melody sub-database can be generated according to the online songs, thereby improving the lyrics and melody coverage and diversity of the song database, thereby improving the professionalism and efficiency of subsequent song generation.

[0123] In another optional embodiment, the music generation method based on intelligent speech analysis may further include the following operations:

[0124] Determining story information corresponding to the voice content, the story information including object information of at least one story object;

[0125] Acquire at least one target story object selected by the user in each story object, and determine a first song optimization parameter according to the object information of each target story object;

[0126] Determine song playing scene information corresponding to the target song, and determine the second song optimization parameter according to the song playing scene information, the song playing scene information including scene type and scene atmosphere;

[0127] The target song is optimized according to the first song optimization parameter and the second song optimization parameter to obtain an optimized target song.

[0128] In this optional embodiment, optionally, the voice content may include story content told by a user, and the story information corresponding to the voice content may include object information of at least one story object, the story object may include at least one of a story character, a story object, etc., and the object information may include one or more of object appearance information, object background information, object personality information, object goal and motivation information, object skill and expertise information, object interpersonal relationship information, object emotional change information, and object growth and change information, etc., which is not limited to this embodiment.

[0129] In this optional embodiment, optionally, the first song optimization parameter is used to optimize the content in the song corresponding to at least one target story object selected by the user, such as optimizing one or more of the proportion of lyrics description of at least one target story object selected by the user, the content described in the lyrics, the emotions expressed by the melody, and the ups and downs of the melody, etc., which is not limited in this embodiment.

[0130] In this optional embodiment, optionally, the song playback scene information may include scene type and scene atmosphere. Taking the user as a child as an example, the scene type may include one or more of a family playback scene type, a school class display playback scene type, a school commendation meeting scene type, a social public display scene type, etc. The scene atmosphere may include one or more of joy, ease, solemnity, seriousness, etc. The second song optimization parameter is used to optimize the lyrics and melody of the target song according to the song playback scene information corresponding to the target song, and the target song is optimized according to the first song optimization parameter and the second song optimization parameter to obtain an optimized target song. This embodiment is not limited thereto.

[0131] It can be seen that implementing this optional embodiment can determine the story information corresponding to the voice content, the story information includes the object information of at least one story object, obtain at least one target story object selected by the user in each story object, and determine the first song optimization parameter according to the object information of each target story object, determine the song playback scene information corresponding to the target song, and determine the second song optimization parameter according to the song playback scene information, the song playback scene information includes the scene type and the scene atmosphere, and optimize the target song according to the first song optimization parameter and the second song optimization parameter to obtain the optimized target song. The target song can be optimized according to the story object narrated by the user and the song playback scene, thereby improving the fit between the target song and the story object and the playback scene, improving the song quality of the generated target song, and thereby improving the user experience.

[0132] Embodiment 2

[0133] See also Figure 2 , Figure 2 : is a flowchart of a music generation method based on intelligent speech analysis disclosed in an embodiment of the present invention. Figure 2 The described music generation method based on intelligent speech analysis can be applied to a music generation device based on intelligent speech analysis, wherein the music generation device based on intelligent speech analysis may include a music generation device for analyzing speech and generating music, and the music generation device may include a display screen, a button, and an intelligent server or intelligent platform for analyzing speech and generating music, and the intelligent server includes a local server or a cloud server, which is not limited in the embodiments of the present invention. Figure 2 As shown, the music generation method based on intelligent speech analysis may include the following operations:

[0134] 201. Obtain a target voice input by a user, and perform voice analysis on the target voice to obtain voice information corresponding to the target voice, where the voice information includes voice content and voice emotion.

[0135] 202. Extract at least one keyword from the voice content according to a preset content extraction rule.

[0136] 203. Analyze the speech emotion to obtain at least one emotion label corresponding to the speech emotion.

[0137] In the embodiment of the present invention, for other descriptions of step 201 to step 203, please refer to the detailed description of step 101 to step 103 in the first embodiment, and the embodiment of the present invention will not be repeated here.

[0138] 204. For each keyword, determine the lyrics classification information corresponding to the keyword according to the lyrics classification rule, and determine at least one lyrics sub-database corresponding to the keyword according to the lyrics classification information.

[0139] In an embodiment of the present invention, optionally, the lyrics classification information corresponding to a keyword may indicate the classification of the lyrics of the keyword, which may specifically include one or more of narrative lyrics, narrative lyrics, descriptive lyrics, reasoning lyrics, and duet lyrics, etc. At least one lyrics sub-database corresponding to the keyword may be determined based on the lyrics classification information.

[0140] 205. For each keyword, in each lyric sub-database corresponding to the keyword, filter at least one lyric segment corresponding to the keyword.

[0141] In an embodiment of the present invention, optionally, at least one lyric fragment corresponding to the keyword can be screened in each lyric sub-database corresponding to the keyword. Specifically, for example, when the keyword corresponds to three lyric sub-databases, at least one lyric fragment can be screened in each of the three lyric sub-databases, or at least one lyric fragment can be screened in one or two of the three lyric sub-databases. The present invention is not limited to this.

[0142] 206. For each emotion tag, determine at least one melody emotion tag corresponding to the emotion tag, and according to each melody emotion tag, determine at least one melody sub-database corresponding to the emotion tag.

[0143] In an embodiment of the present invention, optionally, at least one melody emotion tag corresponding to an emotion tag may include one or more of advancing, rising, masculine, fierce, lyrical and euphemistic, lingering, jumping, exciting, progressive, smooth, cold, sad, etc., and at least one melody sub-database corresponding to each melody emotion tag can be determined based on the emotion tag.

[0144] 207. For each emotion tag, in each melody sub-database corresponding to the emotion tag, filter at least one melody segment corresponding to the emotion tag.

[0145] In an embodiment of the present invention, optionally, at least one melody segment corresponding to the emotion tag can be screened in each melody sub-database corresponding to the emotion tag. Specifically, for example, when the emotion tag corresponds to three melody sub-databases, at least one melody segment can be screened in each of the three melody sub-databases, or at least one lyric segment can be screened in one or two of the three melody sub-databases, and the present invention is not limited thereto.

[0146] 208. Generate a target song corresponding to the target voice according to each lyric segment and each melody segment.

[0147] It can be seen that implementation Figure 2The described music generation method based on intelligent speech analysis can obtain the target speech input by the user, and perform speech analysis on the target speech to obtain the speech content and speech emotion corresponding to the target speech, extract at least one keyword in the speech content according to the preset content extraction rule, analyze the speech emotion, and obtain at least one emotion tag corresponding to the speech emotion, for each keyword, determine the lyrics classification information corresponding to the keyword according to the lyrics classification rule, and determine at least one lyrics sub-database corresponding to the keyword according to the lyrics classification information, for each keyword, in each lyrics sub-database corresponding to the keyword, filter at least one lyrics fragment corresponding to the keyword, for each emotion tag, determine at least one melody emotion tag corresponding to the emotion tag, and determine the melody emotion tag corresponding to the emotion according to each melody emotion tag. At least one melody sub-database corresponding to the tag, for each emotional tag, in each melody sub-database corresponding to the emotional tag, at least one melody fragment corresponding to the emotional tag is screened, and a target song corresponding to the target voice is generated according to each lyrics fragment and each melody fragment. Lyrics fragments and melody fragments that match the content and emotions input by the user can be screened in the database to improve the matching degree between the lyrics fragment and the melody fragment and the content and emotions narrated by the user. A target song corresponding to the target voice can be generated according to each lyrics fragment and each melody fragment, and the matching degree between the generated target song and the content and emotions input by the user is improved, the difficulty of song generation is reduced, and the fun of song generation is improved, so that users can create unique music in the narration process, increase users' interest in music, and enhance users' usage experience.

[0148] In an optional embodiment, generating a target song corresponding to the target voice according to each lyric segment and each melody segment may include the following operations:

[0149] Combining each lyric segment to generate at least one combined lyric, and combining each melody segment to generate at least one combined melody;

[0150] For each combination of lyrics, the combination of lyrics is matched with each combination of melodies to obtain a combination song of the combination of lyrics and each combination of melodies, and a song score of each combination song is determined;

[0151] The song creation habit information corresponding to the user is obtained, and according to the song creation habit information and the song score of each combination of songs, a target song corresponding to the target voice is determined in each combination of songs.

[0152] In this optional embodiment, optionally, combining lyrics for each lyric fragment to generate at least one combination of lyrics may include directly combining the lyrics for each lyric fragment in different orders to generate at least one combination of lyrics, and may also include combining the lyrics for each lyric fragment in different orders, and then intelligently optimizing each combination of lyrics, such as optimizing the words in the combined lyrics, unifying the descriptions of different lyrics before and after, and then obtaining at least one combination of lyrics. It may also include combining the lyrics for each lyric fragment in different orders, and then generating new lyrics corresponding to each combination according to the lyric parameters set by the user. The lyric parameters may include one or more of lyric preference parameters, lyric segmentation parameters, lyric length parameters, etc., which are not limited in this embodiment.

[0153] In this optional embodiment, optionally, combining each melody fragment to generate at least one combined melody may include directly performing melody combinations of each melody fragment in different orders to generate at least one combined melody, or may include combining each melody fragment and then performing intelligent optimization on each combined melody, such as optimizing the melody transitions and melody notes in the combined melody to obtain at least one combined melody. It may also include combining each melody fragment and then generating a new melody corresponding to each combination according to melody parameters set by the user, wherein the melody parameters may include one or more of melody preference parameters, melody segmentation parameters, melody length parameters, etc., which are not limited in this embodiment.

[0154] In this optional embodiment, optionally, for each combination of lyrics, the combination of lyrics is matched with each combination of melody to obtain a combination song of the combination of lyrics and each combination of melody, and the song score of each combination song is determined. Specifically, the lyrics score and melody score of the combination song can be determined separately, and then the song score of the combination song is determined based on the lyrics score and the melody score. The combination song can also be directly analyzed to determine the song score of the combination song. The song creation habit information corresponding to the user may include one or more of song creation length habits, song creation type habits, song creation emotional preference habits, song creation melody habits, song creation lyrics habits, etc. According to the song creation habit information and the song score of each combination song, determining the target song corresponding to the target voice in each combination song may include determining a combination song with a higher song score and a higher match with the song creation habit information in each combination song as the target song corresponding to the target voice, which is not limited in this embodiment.

[0155] It can be seen that the implementation of this optional embodiment can combine lyrics for each lyric fragment to generate at least one combination of lyrics, and combine each melody fragment to generate at least one combination of melody. For each combination of lyrics, the combination of lyrics is matched with each combination of melody to obtain a combination song of the combination of lyrics and each combination of melody, and the song score of each combination of song is determined, and the song creation habit information corresponding to the user is obtained. According to the song creation habit information and the song score of each combination of song, the target song corresponding to the target voice is determined in each combination of song. It is possible to combine multiple lyrics and melodies to generate multiple combinations of lyrics and combination melodies, and determine the target song corresponding to the target voice according to user habits and scores, thereby improving the song creation quality and efficiency of the target song, reducing the difficulty of song generation, and at the same time improving the degree of fit between the target song and user habits, thereby enhancing the user experience and the interest of song creation.

[0156] In another optional embodiment, the music generation method based on intelligent speech analysis may further include the following operations:

[0157] Analyze the originality of the target song to obtain the originality score of the target song, and determine whether the originality score is less than a preset originality score threshold to obtain a determination result;

[0158] When the judgment result indicates that the originality score is less than the originality score threshold, the infringement risk of the target song is determined, and the infringement risk includes the lyrics infringement risk of each lyrics segment and / or the melody infringement risk of each melody segment;

[0159] According to the infringement risk, determine the song modification plan for the target song, which includes a lyrics modification plan and / or a melody modification plan.

[0160] In this optional embodiment, optionally, an originality analysis is performed on the target song to obtain a lyrics originality score, a melody originality score and a comprehensive originality score of the target song. The originality score of the target song may represent a comprehensive originality score. When the judgment result indicates that the originality score is less than the originality score threshold, the infringement risk of the target song is determined. The infringement risk may include the lyrics infringement risk of each lyrics segment and / or the melody infringement risk of each melody segment. When the judgment result indicates that the originality score is greater than the originality score threshold, it may indicate that the originality of the target song is high. According to the infringement risk, a song modification plan for the target song is determined. The song modification plan includes a lyrics modification plan and / or a melody modification plan, which is not limited in this embodiment.

[0161] It can be seen that implementing this optional embodiment can perform originality analysis on the target song, obtain the originality score of the target song, and determine whether the originality score is less than a preset originality score threshold to obtain a judgment result. When the judgment result indicates that the originality score is less than the originality score threshold, the infringement risk of the target song is determined, and based on the infringement risk, a song modification plan for the target song is determined, which can reduce the infringement risk of the target song, reduce the user's infringement probability, and improve the user's creative security and experience.

[0162] In yet another optional embodiment, determining a song modification scheme for a target song according to the infringement risk may include the following operations:

[0163] For each lyric segment, when the lyric infringement risk of the lyric segment is greater than a preset lyric infringement risk threshold, the lyric segment is determined to be a risky lyric segment;

[0164] For each risky lyrics segment, determine the lyrics copyright information of the risky lyrics segment, and judge whether the lyrics copyright information of the risky lyrics segment meets the lyrics copyright acquisition conditions of the user; when the lyrics copyright information of the risky lyrics segment meets the lyrics copyright acquisition conditions of the user, determine a lyrics copyright acquisition plan for the risky lyrics segment; when the lyrics copyright information of the risky lyrics segment does not meet the lyrics copyright acquisition conditions of the user, determine at least one replacement lyrics segment corresponding to the risky lyrics segment in the song database, and determine a lyrics replacement plan for the risky lyrics segment based on each replacement lyrics segment;

[0165] Determine a lyrics modification plan based on the lyrics copyright acquisition plan or lyrics replacement plan for each risky lyrics segment; and / or,

[0166] For each melody segment, when the melody infringement risk of the melody segment is greater than a preset melody infringement risk threshold, the melody segment is determined to be a risky melody segment;

[0167] For each risky melody segment, determine the melody copyright information of the risky melody segment, and judge whether the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user; when the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user, determine a melody copyright acquisition scheme for the risky melody segment; when the melody copyright information of the risky melody segment does not meet the melody copyright acquisition condition of the user, determine at least one replacement melody segment corresponding to the risky melody segment in a song database, and determine a melody replacement scheme for the risky melody segment according to each replacement melody segment;

[0168] Determine a melody modification plan based on a melody copyright acquisition plan or a melody replacement plan for each risky melody segment;

[0169] A song modification scheme for the target song is determined according to the lyrics modification scheme and / or the melody modification scheme.

[0170] In this optional embodiment, it is optional to determine whether the risk of lyric infringement of each lyric segment is greater than a preset lyric infringement risk threshold. When the risk of lyric infringement of the lyric segment is greater than the preset lyric infringement risk threshold, the lyric segment is determined to be a risky lyric segment, and the lyric copyright information of each risky lyric segment is determined. The lyric copyright information may include one or more of the copyright owner information, copyright acquisition path information, copyright acquisition condition information, etc. of the risky lyric segment. The user's lyric copyright acquisition condition may include the maximum condition set by the user for acquiring the lyric copyright, such as a maximum payment amount, etc. When the lyric copyright information of the risky lyric segment meets the user's lyric copyright acquisition condition, a lyric copyright acquisition plan for the risky lyric segment is determined, such as a copyright purchase plan. When the lyric copyright information of the risky lyric segment does not meet the user's lyric copyright acquisition condition, at least one replacement lyric segment corresponding to the risky lyric segment is determined in the song database, and a lyric replacement plan for the risky lyric segment is determined based on each replacement lyric segment. This embodiment does not limit this.

[0171] In this optional embodiment, it is optional to determine whether the melody infringement risk of each melody segment is greater than a preset melody infringement risk threshold. When the melody infringement risk of the melody segment is greater than the preset melody infringement risk threshold, the melody segment is determined to be a risky melody segment, and the melody copyright information of each risky melody segment is determined. The melody copyright information may include one or more of the copyright owner information, copyright acquisition path information, copyright acquisition condition information, etc. of the risky melody segment.

[0172] The user's melody copyright acquisition conditions may include the maximum conditions for acquiring the melody copyright set by the user, such as a maximum payment amount, etc. When the melody copyright information of the risky melody fragment meets the user's melody copyright acquisition conditions, a melody copyright acquisition plan for the risky melody fragment is determined, such as a copyright purchase plan. When the melody copyright information of the risky melody fragment does not meet the user's melody copyright acquisition conditions, at least one replacement melody fragment corresponding to the risky melody fragment is determined in the song database, and a melody replacement plan for the risky melody fragment is determined based on each replacement melody fragment. This embodiment does not limit this.

[0173] It can be seen that the implementation of this optional embodiment can obtain the copyright of lyrics and melody based on the copyright information of the song when the copyright acquisition conditions are met, and modify the song and melody when the copyright acquisition conditions are not met. It can provide users with a variety of modification plans, reduce the infringement risk of the target song, reduce the user's infringement probability, and improve the user's creation security and experience.

[0174] Embodiment 3

[0175] See also Figure 3 , Figure 3 : is a schematic diagram of a music generation device based on intelligent speech analysis disclosed in an embodiment of the present invention. Figure 3 The described music generation apparatus based on intelligent voice analysis may include a music generation device for analyzing voice and generating music, and the music generation device may include a display screen, a button, an intelligent server or an intelligent platform for analyzing voice and generating music, and the intelligent server may include a local server or a cloud server, which is not limited in the embodiments of the present invention. Figure 3 As shown, the music generation device based on intelligent voice analysis may include:

[0176] The acquisition module 301 is used to acquire the target speech input by the user, and perform speech analysis on the target speech to obtain speech information corresponding to the target speech, where the speech information includes speech content and speech emotion;

[0177] An extraction module 302, configured to extract at least one keyword from the speech content according to a preset content extraction rule;

[0178] An analysis module 303 is used to analyze the speech emotion and obtain at least one emotion label corresponding to the speech emotion;

[0179] The generation module 304 is used to generate a target song corresponding to the target speech according to each keyword and each emotion tag.

[0180] It can be seen that implementation Figure 3 The described music generation device based on intelligent speech analysis can obtain the target speech input by the user, and perform speech analysis on the target speech to obtain the speech content and speech emotion corresponding to the target speech, extract at least one keyword in the speech content according to preset content extraction rules, analyze the speech emotion, and obtain at least one emotion tag corresponding to the speech emotion, generate a target song corresponding to the target speech according to each keyword and each emotion tag, and can generate corresponding songs according to the content and emotions narrated by the user, reduce the difficulty of song generation, and improve the fun of song generation, so that the user can create unique music in the narration process, increase the user's interest in music, and enhance the user experience.

[0181] In an optional embodiment, if Figure 4 As shown, the acquisition module 301 is also used to acquire a network song collection, which includes a network lyrics collection and a network melody collection;

[0182] The music generation device based on intelligent voice analysis may also include:

[0183] The classification module 305 is used to classify the lyrics of the network lyrics set according to the preset lyrics classification rules to obtain at least one lyrics sub-database corresponding to the network lyrics set;

[0184] The analysis module 303 is further used to perform melody analysis on each network melody in the network melody set to determine at least one melody emotion label corresponding to each network melody;

[0185] The classification module 305 is further used to classify each network melody according to each melody emotion label corresponding to the network melody, and obtain at least one melody sub-database corresponding to the network melody set;

[0186] The generating module 304 is further used to generate a song database according to each lyrics sub-database and each melody sub-database.

[0187] It can be seen that implementation Figure 4 The described music generation device based on intelligent voice analysis can obtain a collection of online songs, which includes a collection of online lyrics and a collection of online melody. According to preset lyrics classification rules, the lyrics of the online lyrics collection are classified to obtain at least one lyrics sub-database corresponding to the online lyrics collection. Melody analysis is performed on each online melody in the online melody collection to determine at least one melody emotion tag corresponding to each online melody. For each online melody, the melody is classified according to each melody emotion tag corresponding to the online melody to obtain at least one melody sub-database corresponding to the online melody collection. A song database is generated according to each lyrics sub-database and each melody sub-database. The corresponding lyrics sub-database and melody sub-database can be generated according to the online songs, thereby improving the lyrics and melody coverage and diversity of the song database, thereby improving the professionalism and efficiency of subsequent song generation.

[0188] In another optional embodiment, Figure 4 As shown, the target song includes target lyrics and target melody;

[0189] The specific method in which the generating module 304 generates a target song corresponding to the target speech according to each keyword and each emotion tag includes:

[0190] For each keyword, determining lyrics classification information corresponding to the keyword according to the lyrics classification rule, and determining at least one lyrics sub-database corresponding to the keyword according to the lyrics classification information;

[0191] For each keyword, in each lyrics sub-database corresponding to the keyword, selecting at least one lyrics segment corresponding to the keyword;

[0192] For each emotion tag, determining at least one melody emotion tag corresponding to the emotion tag, and according to each melody emotion tag, determining at least one melody sub-database corresponding to the emotion tag;

[0193] For each emotion tag, in each melody sub-database corresponding to the emotion tag, selecting at least one melody segment corresponding to the emotion tag;

[0194] According to each lyric segment and each melody segment, a target song corresponding to the target voice is generated.

[0195] It can be seen that implementation Figure 4 The described music generation device based on intelligent speech analysis can obtain the target speech input by the user, and perform speech analysis on the target speech to obtain the speech content and speech emotion corresponding to the target speech, extract at least one keyword in the speech content according to the preset content extraction rule, analyze the speech emotion, and obtain at least one emotion tag corresponding to the speech emotion, for each keyword, determine the lyrics classification information corresponding to the keyword according to the lyrics classification rule, and determine at least one lyrics sub-database corresponding to the keyword according to the lyrics classification information, for each keyword, in each lyrics sub-database corresponding to the keyword, filter at least one lyrics fragment corresponding to the keyword, for each emotion tag, determine at least one melody emotion tag corresponding to the emotion tag, and determine the melody emotion tag corresponding to the emotion according to each melody emotion tag. At least one melody sub-database corresponding to the tag, for each emotional tag, in each melody sub-database corresponding to the emotional tag, at least one melody fragment corresponding to the emotional tag is screened, and a target song corresponding to the target voice is generated according to each lyrics fragment and each melody fragment. Lyrics fragments and melody fragments that match the content and emotions input by the user can be screened in the database to improve the matching degree between the lyrics fragment and the melody fragment and the content and emotions narrated by the user. A target song corresponding to the target voice can be generated according to each lyrics fragment and each melody fragment, and the matching degree between the generated target song and the content and emotions input by the user is improved, the difficulty of song generation is reduced, and the fun of song generation is improved, so that users can create unique music in the narration process, increase users' interest in music, and enhance users' usage experience.

[0196] In yet another optional embodiment, Figure 4 As shown, the specific method of generating the target song corresponding to the target voice according to each lyric segment and each melody segment by the generating module 304 includes:

[0197] Combining each lyric segment to generate at least one combined lyric, and combining each melody segment to generate at least one combined melody;

[0198] For each combination of lyrics, the combination of lyrics is matched with each combination of melodies to obtain a combination song of the combination of lyrics and each combination of melodies, and a song score of each combination song is determined;

[0199] The song creation habit information corresponding to the user is obtained, and according to the song creation habit information and the song score of each combination of songs, a target song corresponding to the target voice is determined in each combination of songs.

[0200] It can be seen that implementation Figure 4 The described music generation device based on intelligent voice analysis can combine lyrics for each lyric fragment to generate at least one combination of lyrics, and combine each melody fragment to generate at least one combination of melody. For each combination of lyrics, the combined lyrics are matched with each combination of melody to obtain a combination song of the combined lyrics and each combination of melody, and the song score of each combination of song is determined, and the song creation habit information corresponding to the user is obtained. According to the song creation habit information and the song score of each combination of song, the target song corresponding to the target voice is determined in each combination of song. It is possible to combine multiple lyrics and melodies to generate multiple combinations of lyrics and combination melodies, and the target song corresponding to the target voice is determined according to the user habits and scores, thereby improving the song creation quality and efficiency of the target song, reducing the difficulty of song generation, and at the same time improving the degree of fit between the target song and the user habits, thereby enhancing the user experience and the interest of song creation.

[0201] In yet another optional embodiment, Figure 4 As shown, the music generation device based on intelligent voice analysis may also include:

[0202] A first determination module 306, configured to determine story information corresponding to the voice content, the story information including object information of at least one story object;

[0203] The acquisition module 301 is further used to acquire at least one target story object selected by the user in each story object, and determine the first song optimization parameter according to the object information of each target story object;

[0204] The first determination module 306 is further used to determine the song playing scene information corresponding to the target song, and determine the second song optimization parameter according to the song playing scene information, the song playing scene information includes the scene type and the scene atmosphere;

[0205] The optimization module 307 is used to optimize the target song according to the first song optimization parameter and the second song optimization parameter to obtain the optimized target song.

[0206] It can be seen that implementation Figure 4The described music generation device based on intelligent speech analysis can determine the story information corresponding to the speech content, the story information includes the object information of at least one story object, obtain at least one target story object selected by the user in each story object, and determine the first song optimization parameter according to the object information of each target story object, determine the song playback scene information corresponding to the target song, and determine the second song optimization parameter according to the song playback scene information, the song playback scene information includes the scene type and the scene atmosphere, and optimize the target song according to the first song optimization parameter and the second song optimization parameter to obtain the optimized target song. The target song can be optimized according to the story object told by the user and the song playback scene, thereby improving the fit between the target song and the story object and the playback scene, improving the song quality of the generated target song, and thus improving the user experience.

[0207] In yet another optional embodiment, Figure 4 As shown, the analysis module 303 is also used to analyze the originality of the target song, obtain the originality score of the target song, and determine whether the originality score is less than a preset originality score threshold to obtain a determination result;

[0208] The music generation device based on intelligent voice analysis may also include:

[0209] The second determination module 308 is used to determine the infringement risk of the target song when the judgment result indicates that the originality score is less than the originality score threshold, the infringement risk including the lyrics infringement risk of each lyrics segment and / or the melody infringement risk of each melody segment;

[0210] The second determination module 308 is further used to determine a song modification plan for the target song according to the infringement risk, and the song modification plan includes a lyrics modification plan and / or a melody modification plan.

[0211] It can be seen that implementation Figure 4 The described music generation device based on intelligent voice analysis can perform originality analysis on a target song, obtain an originality score for the target song, and determine whether the originality score is less than a preset originality score threshold to obtain a judgment result. When the judgment result indicates that the originality score is less than the originality score threshold, the infringement risk of the target song is determined. Based on the infringement risk, a song modification plan for the target song is determined, which can reduce the infringement risk of the target song, reduce the user's chance of infringement, and improve the user's creative security and experience.

[0212] In yet another optional embodiment, Figure 4 As shown, the second determination module 308 determines the specific manner of the song modification scheme for the target song according to the infringement risk, including:

[0213] For each lyric segment, when the lyric infringement risk of the lyric segment is greater than a preset lyric infringement risk threshold, the lyric segment is determined to be a risky lyric segment;

[0214] For each risky lyrics segment, determine the lyrics copyright information of the risky lyrics segment, and judge whether the lyrics copyright information of the risky lyrics segment meets the lyrics copyright acquisition conditions of the user; when the lyrics copyright information of the risky lyrics segment meets the lyrics copyright acquisition conditions of the user, determine a lyrics copyright acquisition plan for the risky lyrics segment; when the lyrics copyright information of the risky lyrics segment does not meet the lyrics copyright acquisition conditions of the user, determine at least one replacement lyrics segment corresponding to the risky lyrics segment in the song database, and determine a lyrics replacement plan for the risky lyrics segment based on each replacement lyrics segment;

[0215] Determine a lyrics modification plan based on the lyrics copyright acquisition plan or lyrics replacement plan for each risky lyrics segment; and / or,

[0216] For each melody segment, when the melody infringement risk of the melody segment is greater than a preset melody infringement risk threshold, the melody segment is determined to be a risky melody segment;

[0217] For each risky melody segment, determine the melody copyright information of the risky melody segment, and judge whether the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user; when the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user, determine a melody copyright acquisition scheme for the risky melody segment; when the melody copyright information of the risky melody segment does not meet the melody copyright acquisition condition of the user, determine at least one replacement melody segment corresponding to the risky melody segment in a song database, and determine a melody replacement scheme for the risky melody segment according to each replacement melody segment;

[0218] Determine a melody modification plan based on a melody copyright acquisition plan or a melody replacement plan for each risky melody segment;

[0219] A song modification scheme for the target song is determined according to the lyrics modification scheme and / or the melody modification scheme.

[0220] It can be seen that implementation Figure 4 The described music generation device based on intelligent voice analysis can obtain the copyright of lyrics and melody based on the copyright information of the lyrics and melody of the song when the copyright acquisition conditions are met, and modify the song and melody when the copyright acquisition conditions are not met. It can provide users with multiple modification plans, reduce the infringement risk of the target song, reduce the user's infringement probability, and improve the user's creation security and experience.

[0221] Embodiment 4

[0222] See also Figure 5 , Figure 5 FIG. 1 is a schematic diagram of the structure of another music generation device based on intelligent speech analysis disclosed in an embodiment of the present invention. Figure 5 As shown, the music generation device based on intelligent voice analysis may include:

[0223] A memory 401 storing executable program codes;

[0224] a processor 402 coupled to the memory 401;

[0225] The processor 402 calls the executable program code stored in the memory 401 to execute the steps in the music generation method based on intelligent speech analysis described in the first embodiment of the present invention or the second embodiment of the present invention.

[0226] Embodiment 5

[0227] An embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the steps in the music generation method based on intelligent speech analysis described in Embodiment 1 or Embodiment 2 of the present invention.

[0228] Embodiment 6

[0229] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps in the music generation method based on intelligent speech analysis described in Example 1 or Example 2.

[0230] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0231] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0232] Finally, it should be noted that the music generation method and device based on intelligent speech analysis disclosed in the embodiments of the present invention only disclose the preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A music generation method based on intelligent speech analysis, characterized in that: The method comprises: Acquire a target voice input by a user, and perform voice analysis on the target voice to obtain voice information corresponding to the target voice, wherein the voice information includes voice content and voice emotion; Extracting at least one keyword from the speech content according to a preset content extraction rule; Analyze the speech emotion to obtain at least one emotion label corresponding to the speech emotion; According to each of the keywords and each of the emotion tags, a target song corresponding to the target speech is generated.

2. The music generation method based on intelligent speech analysis according to claim 1, characterized in that: The method further comprises: Acquire a network song collection, wherein the network song collection includes a network lyrics collection and a network melody collection; Classifying the network lyrics collection according to a preset lyrics classification rule to obtain at least one lyrics sub-database corresponding to the network lyrics collection; Performing melody analysis on each network melody in the network melody set to determine at least one melody emotion label corresponding to each network melody; For each of the network melodies, classifying the network melody according to each of the melody emotion tags corresponding to the network melody, and obtaining at least one melody sub-database corresponding to the network melody set; A song database is generated based on each of the lyrics sub-databases and each of the melody sub-databases.

3. The music generation method based on intelligent speech analysis according to claim 2 is characterized in that: The target song includes target lyrics and target melody; The step of generating a target song corresponding to the target speech according to each of the keywords and each of the emotion tags comprises: For each of the keywords, determining the lyrics classification information corresponding to the keyword according to the lyrics classification rule, and determining at least one of the lyrics sub-databases corresponding to the keyword according to the lyrics classification information; For each of the keywords, in each of the lyrics sub-databases corresponding to the keyword, screening at least one lyrics segment corresponding to the keyword; For each of the emotion tags, determining at least one of the melody emotion tags corresponding to the emotion tag, and according to each of the melody emotion tags, determining at least one of the melody sub-databases corresponding to the emotion tag; For each of the emotion tags, in each of the melody sub-databases corresponding to the emotion tag, screening at least one melody segment corresponding to the emotion tag; A target song corresponding to the target voice is generated according to each of the lyrics segments and each of the melody segments.

4. The music generation method based on intelligent speech analysis according to claim 3 is characterized in that: The step of generating a target song corresponding to the target voice according to each of the lyrics segments and each of the melody segments comprises: Combining each of the lyrics fragments to generate at least one combined lyric, and combining each of the melody fragments to generate at least one combined melody; For each of the combined lyrics, the combined lyrics are matched with each of the combined melodies to obtain a combined song of the combined lyrics and each of the combined melodies, and a song score of each of the combined songs is determined; The song creation habit information corresponding to the user is obtained, and according to the song creation habit information and the song score of each of the combination songs, a target song corresponding to the target voice is determined in each of the combination songs.

5. The music generation method based on intelligent speech analysis according to any one of claims 1 to 4, characterized in that: The method further comprises: Determining story information corresponding to the voice content, the story information including object information of at least one story object; Acquire at least one target story object selected by the user in each of the story objects, and determine a first song optimization parameter according to the object information of each of the target story objects; Determine song playing scene information corresponding to the target song, and determine a second song optimization parameter according to the song playing scene information, wherein the song playing scene information includes a scene type and a scene atmosphere; The target song is optimized according to the first song optimization parameter and the second song optimization parameter to obtain an optimized target song.

6. The music generation method based on intelligent speech analysis according to claim 3 or 4, characterized in that: The method further comprises: Performing originality analysis on the target song to obtain an originality score of the target song, and determining whether the originality score is less than a preset originality score threshold to obtain a determination result; When the judgment result indicates that the originality score is less than the originality score threshold, determining the infringement risk of the target song, the infringement risk including the lyrics infringement risk of each of the lyrics segments and / or the melody infringement risk of each of the melody segments; According to the infringement risk, a song modification plan for the target song is determined, and the song modification plan includes a lyrics modification plan and / or a melody modification plan.

7. The music generation method based on intelligent speech analysis according to claim 6 is characterized in that: Determining a song modification plan for the target song based on the infringement risk includes: For each of the lyrics segments, when the lyrics infringement risk of the lyrics segment is greater than a preset lyrics infringement risk threshold, the lyrics segment is determined to be a risky lyrics segment; For each of the risky lyrics fragments, determine the lyrics copyright information of the risky lyrics fragment, and judge whether the lyrics copyright information of the risky lyrics fragment meets the lyrics copyright acquisition conditions of the user; when the lyrics copyright information of the risky lyrics fragment meets the lyrics copyright acquisition conditions of the user, determine a lyrics copyright acquisition plan for the risky lyrics fragment; when the lyrics copyright information of the risky lyrics fragment does not meet the lyrics copyright acquisition conditions of the user, determine at least one replacement lyrics fragment corresponding to the risky lyrics fragment in the song database, and determine a lyrics replacement plan for the risky lyrics fragment based on each of the replacement lyrics fragments; Determine a lyrics modification plan based on a lyrics copyright acquisition plan or a lyrics replacement plan for each of the risky lyrics segments; and / or, For each of the melody segments, when the melody infringement risk of the melody segment is greater than a preset melody infringement risk threshold, the melody segment is determined to be a risky melody segment; For each of the risky melody segments, determine the melody copyright information of the risky melody segment, and judge whether the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user; when the melody copyright information of the risky melody segment meets the melody copyright acquisition condition of the user, determine a melody copyright acquisition scheme for the risky melody segment; when the melody copyright information of the risky melody segment does not meet the melody copyright acquisition condition of the user, determine at least one replacement melody segment corresponding to the risky melody segment in the song database, and determine a melody replacement scheme for the risky melody segment according to each of the replacement melody segments; Determine a melody modification plan according to a melody copyright acquisition plan or a melody replacement plan of each of the risky melody fragments; A song modification scheme for the target song is determined according to the lyrics modification scheme and / or the melody modification scheme.

8. A music generation device based on intelligent speech analysis, characterized in that: The device comprises: An acquisition module is used to acquire a target voice input by a user, and perform voice analysis on the target voice to obtain voice information corresponding to the target voice, wherein the voice information includes voice content and voice emotion; An extraction module, used to extract at least one keyword from the speech content according to a preset content extraction rule; An analysis module, used to analyze the speech emotion and obtain at least one emotion label corresponding to the speech emotion; A generation module is used to generate a target song corresponding to the target speech according to each of the keywords and each of the emotion tags.

9. A music generation device based on intelligent speech analysis, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the music generation method based on intelligent speech analysis as described in any one of claims 1-7.

10. A computer storage medium, characterized in that: The computer storage medium stores computer instructions, which, when called, are used to execute the music generation method based on intelligent speech analysis as described in any one of claims 1-7.

Images