Artificial intelligence assisted music editing method, apparatus and device, and storage medium

Abstract

The invention belongs to the field of artificial intelligence, and discloses a method for assisting music editing by artificial intelligence, which comprises the following steps of: constructing a music database based on basic structure information of music, including a music hierarchical structure, music sentences containing lyrics, music segments, corresponding lyric text-timestamp data pairs, and one or more previous interplay segment starting and ending timestamps; the selection range of the selected music clip is determined, and the terminal identifies lyric word-by-word text-timestamp data and / or previous interplay music data according to the selected music clip; taking the corresponding stored phrases and / or music segments as selected segments, and taking the stored previous interplay segments as selected previous interplay segments; and outputting the target music. The invention further discloses a corresponding system and device and a storage medium. According to the method, the basic structure information of the music is applied in the process of assisting in music editing through artificial intelligence, so that the edited music clip is more in line with potential music image emotion expected to be expressed by a user (especially a non-music professional user).

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and in particular to a method, device, equipment and storage medium for artificial intelligence-assisted music editing. Background Art

[0002] Short video programs are short videos shot by users as a sharing platform for online content. In short video programs, users have the need to select music clips.

[0003] A music editing interface is provided on the short video program. The music editing interface has a music display control, which usually includes one or more of a music progress bar, a music spectrogram, and a music lyrics graph. The music editing interface has a segment selection control (such as a vertical line or a round needle indicator). By positioning the selection control at the position pointed to by the music playback progress bar, the music spectrogram, and the music lyrics graph, the short video program determines the selected music segment according to the position pointed to by the segment selection control and the music segment selection rules of the short video program, such as preset recommended music segments. Since (a) users usually select the desired music segment based on the lyrics segment, selecting through the playback progress bar, music spectrogram, etc. does not conform to user habits and increases the user's usage cost, (b) the playback progress bar and the music spectrogram are limited in length, and the accuracy of the user's finger sliding operation is low, making it difficult to accurately locate a certain second or millisecond in a piece of music, and it is difficult for the user to accurately select the desired music segment, and (c) the preset music segment editing scheme as a candidate recommendation scheme cannot cover all the music segments that users expect to select. Therefore, the ease of use, efficiency, and flexibility of the current technical solution for editing music segments need to be improved. Summary of the invention

[0004] In view of the problems in the prior art, the present invention provides a method, device, equipment and storage medium for artificial intelligence-assisted music editing.

[0005] In order to achieve the above purpose, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a method for artificial intelligence-assisted music editing, comprising the following steps: Based on the basic structural information of the music, a music database is constructed, wherein the music database includes: a music hierarchy structure, i.e., a permutation and combination hierarchy between phrases, sections, and pre-interlude segments, a phrase, section containing lyrics, and corresponding lyrics word-by-word text-timestamp data pairs, and one or more pre-interlude segment start and end timestamps; Determine the selection range of the selected music segment, the terminal identifies the lyrics verbatim text-time stamp data and / or pre-interlude music data of the stored phrases and / or music segments in the database corresponding to the selected music segment; and uses the corresponding stored phrases and / or music segments as the selected segment, and the stored pre-interlude segment as the selected pre-interlude segment; Output the target music, the target music contains one or more selected phrases, or, a musical phrase and / or one or more prelude fragments, Or, multiple sections and / or one or more prelude segments.

[0006] In the method of artificial intelligence-assisted music editing, further, a music segment contains multiple phrases.

[0007] In the method of artificial intelligence assisted music editing, further, the lyrics verbatim text-timestamp data includes the start and end points of the lyrics text of the music clip accessed from the database and the corresponding timestamp data pairs.

[0008] In the method of artificial intelligence-assisted music editing, further, the pre-interlude data includes the start and end timestamps of the pre-interlude music clip.

[0009] In the method of artificial intelligence-assisted music editing, further, the method of determining the selected music clip is: by locating the position of the click or double-click operation, one or more phrases or pre-interlude fragments covered by it are identified as the selected music clip; or, by detecting the range of the sliding operation, one or more phrases or pre-interlude fragments covered by it are identified as the selected music clip; or, by parsing the voice instruction, the specified music clip therein is identified as one of the selected music clips.

[0010] In the method of artificial intelligence-assisted music editing, further, when the selected fragment selected by the user for the first time is multiple phrases, and the starting point of the selected fragment is the same as the starting point of the segment or the end point of the selected fragment is the same as the end point of the segment, a prompt is given to select the corresponding segment as the selected fragment. If the user selects no, the corresponding multiple phrases are used as the selected fragment; if the user selects yes, the segment covering the above multiple phrases is reset as the selected fragment; when the reset selected fragment is a segment, and the starting point of the selected fragment is the same as the end point of one or more pre-interlude segments or the end point of the selected fragment is the same as the starting point of one or more pre-interlude segments, a prompt is given to select the pre-interlude selection control. If the user selects no, the corresponding segment is used as the target music for output; if the user selects yes, the segment and the pre-interlude segment are reset as the target music for output.

[0011] For example, a user wants to select a certain segment of a song. For example, the user wants to select a certain segment, and the song is composed of (a) prelude, (b) verse section 1, (c) verse section 2, (d) chorus section 1, (e) chorus section 2, (f) interlude, (g) chorus section 1, (h) verse section 2, (i) chorus section 1, (j) chorus section 2, and (k) epilogue.

[0012] AWhen the user expects to select the two phrases in (h), Right now" OP...Q R......ST", That is, "O" at 2 minutes 53 seconds 568 milliseconds to "T" at 3 minutes 01 seconds 122 milliseconds. The user slides the music lyrics chart to any text of "OP...Q" and any text of "R...ST" in the segment, triggering the terminal to automatically select the music segment of "O P...Q., R...ST" from 2 minutes 53 seconds 568 milliseconds to 3 minutes 01 seconds 122 milliseconds on the play progress bar or the music spectrogram.

[0013] B When the user wants to select multiple phrases of (g) (h), That is, "ABC D.....EFG,......,R......ST", That is, "A" at 1 minute 22 seconds 362 milliseconds to "T" at 3 minutes 01 seconds 122 milliseconds. The user slides the music lyrics chart to any text in "ABC D.....EFG" and any text in "X......YZ" in the segment, triggering the terminal to automatically select the music segment from 1 minute 22 seconds 362 milliseconds to "ABC D...... EFG,......,X......YZ" at 3 minutes 01 seconds 122 milliseconds on the play progress bar or the music spectrogram.

[0014] At this time, in this example, the "R......ST" of the selected music segment is the same as the starting point of the music segment (h). On the music editing page, a prompt is displayed to determine whether to select the corresponding music segment (g) (h) as the selected segment. If the user selects "no", the corresponding multiple phrases will be selected as the selected segment. If the user selects "yes", the music segment covering the above multiple phrases will be reset as the selected segment.

[0015] At this time, if the user selects yes, the (g) (h) music segment will be selected as the selected segment. In this example, there is an interlude segment (f) on the left side of (g). The pre-interlude selection control is displayed on the music editing interface. The user clicks the pre-interlude selection control, triggering the terminal to automatically select the music segment containing the music segment and the interlude from 1 minute 03 seconds 129 milliseconds to 3 minutes 01 seconds 122 milliseconds on the playback progress bar or the music spectrogram as the output target music, namely (f) (g) (h).

[0016] In a second aspect, the present invention provides application of the above-mentioned artificial intelligence assisted music editing method in video editing.

[0017] Furthermore, the application of the method of artificial intelligence assisted music editing in video editing comprises the following steps: Step 1, obtaining the video length of the short video; Step 2, determining the timestamp of the edited music segment and the alignment timestamp of the edited music segment and the short video according to the video length of the short video, the target music duration, the stored music segment timestamp and the pre-interlude timestamp.

[0018] Further, the method for determining the timestamp of the edited music segment includes: The duration of the target music clip is subtracted from the video length of the short video to obtain the subtracted duration.

[0019] When the subtracted duration is less than or equal to 0, the starting point of the edited music clip is set to the starting point of the target music clip, and the end point of the edited music clip is reset to the timestamp of the starting point of the target music clip plus the video length of the short video.

[0020] When the subtracted duration is greater than 0 and less than a threshold, the starting point of the clipped music segment is set as the starting point of the target music segment, and the end point of the clipped music segment is reset to the end point of the target music segment.

[0021] When the subtracted duration is greater than a threshold greater than 0, the method for determining the start and end timestamps of the clipped music segment is: Traverse the duration of the left and right pre-interludes of the target music clip, and the added duration of the left and right pre-interludes, and obtain the clip that is closest to the duration after subtraction as the pre-interlude clip, set the starting point of the edited music clip to the smaller of the starting point of the stored phrase clip and the starting point of the pre-interlude clip, set the starting point of the edited music clip to the smaller of the end point of the selected music clip and the end point of the pre-interlude clip, and set the end point of the edited music clip to the larger end point of the pre-interlude clip and the starting point of the selected music clip.

[0022] The terminal determines the alignment timestamp between the clipped music segment and the video length of the short video according to the video length of the short video and the time length of the clipped music segment.

[0023] In a third aspect, the present invention further provides an artificial intelligence-assisted music editing device, the device comprising: The display module is used to display a music editing interface, wherein the music editing interface includes: a music display control and a segment selection control of the selected music; and a pre-interlude selection control for displaying the pre-interlude music segment of the selected music; An interaction module, configured to receive a selection operation on the segment selection control; and receive a trigger operation on the prelude and interlude selection control; an acquisition module, configured to, upon receiving a selection operation on the segment selection control, obtain a stored phrase segment corresponding to the selected music according to the selection operation; The reset module, when receiving a trigger operation on the pre-interlude selection control, resets the selected music segment to obtain the target music according to the selection operation and the pre-interlude segment.

[0024] In a fourth aspect, the present invention also provides a computer device, comprising: a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement the artificial intelligence assisted music editing method as described in the above aspects.

[0025] In a fifth aspect, the present invention also provides a computer-readable storage medium, wherein the storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement the artificial intelligence assisted music editing method as described in the above aspects.

[0026] Beneficial effects of the present invention: The present application obtains music lyrics text-timestamp data pairs and pre-interlude timestamp data, and uses a combined control process of music display controls, clip selection controls, and pre-interlude selection controls to reduce the difficulty of manually locating music clips when users manually select music clips on target music, and select the selected music clips more flexibly, efficiently and quickly, making music editing fast and efficient.

[0027] The feature of this application is that the basic structural information of the music is applied in the process of artificial intelligence-assisted music editing, so that the edited music clips are more in line with the potential musical image emotions that users (especially non-music professionals) want to express. Specifically, from a music professional perspective, phrases are the basic elements that constitute a musical segment. The combination and arrangement of phrases form the unique style of the musical segment and express the unique image of the musical segment. The connection and transition between the musical segments through the prelude and interlude express the unique style and image of the music. This application automatically identifies the structure of the music and its corresponding timestamps to achieve this, by segmenting the musical segment, phrase and prelude, and obtaining the start and end timestamps of the elements at each level and the word-by-word timestamps of the lyrics of the music to assist in completing music editing. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0029] Figure 1 It is a schematic diagram of the interface of the music editing method provided by the related technology of this application; Figure 2 is a flow chart of an artificial intelligence-assisted music editing method provided by an exemplary embodiment of the present application; Figure 3 This is a schematic diagram of a clip selection control interface of an artificial intelligence-assisted music editing method provided by an exemplary embodiment of the present application; Figure 4 This is a schematic diagram of an interface of a prelude and interlude selection control of an artificial intelligence-assisted music editing method provided by an exemplary embodiment of the present application; Figure 5 It is a schematic diagram of the “music structure” of this application; Figure 6 It is a schematic diagram of an interface of an artificial intelligence-assisted music editing method in a related art provided by another exemplary embodiment of the present application; Figure 7 is a schematic diagram of an interface for artificial intelligence-assisted music editing provided by another exemplary embodiment of the present application; Figure 8 is a flow chart of an artificial intelligence-assisted music editing method provided by another exemplary embodiment of the present application; Fig. 9 is a flow chart of an artificial intelligence-assisted music editing method provided by another exemplary embodiment of the present application; Fig.10It is a flowchart of an artificial intelligence-assisted music editing method provided by another exemplary embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0031] The concepts designed in this article are defined as follows: Selected music: a music clip selected by at least one of clicking, double-clicking, sliding, dragging, pressing, and voice command. Clicking or double-clicking may be selecting a music clip by clicking a line of lyrics of the music.

[0032] Storing music: Building a music database based on the basic structural information of the music, and storing phrases and passages in the database.

[0033] Target music: After the phrase fragments are stored according to the selected music, the output music fragments containing phrases, sections and the selected prelude fragments are formed. The complete music work before the selection, editing and other operations begin.

[0034] Edited music: The music clip of the target music after selection, editing, etc. In the application scenario of short video soundtrack, it is the music clip that matches the length of the video after the target music is edited. The edited music is part of the target music.

[0035] Music editing is the process of cutting and editing music. Music editing includes at least one of selecting, cutting, arranging, rearranging, and superimposing music. This function is widely used in various applications. For example, in short video production or short video sharing applications, it is used to compose music for short videos; in video editing software, it is used to compose music for videos; in ringtone generation software, it is used to capture music clips; in music playback software, it is used to capture music clips and share them with friends; in singing applications, it is used to capture some music clips for singing, etc.

[0036] This application provides an artificial intelligence-assisted music editing method, and briefly introduces the method by taking its application in short video production software as an example.

[0037] The process is as follows Figure 1 As shown in the figure, after a user shoots or uploads a short video, the user can select the background music of the video on the music interface. After selecting a song, the user clicks the music editing control to enter the music editing interface. In the related art, the user selects a music clip by sliding the song's sound spectrum or the playback progress bar on the music editing interface. However, since the sound spectrum or the playback progress bar is very short, it is difficult to accurately locate a certain second or millisecond in a piece of music by sliding the operation to accurately select the desired music clip. Specifically, Figure 1As shown, the user wants to select a certain segment of a certain song, for example, "O P...Q, R...ST" in the target music, that is, "O" at 2 minutes 23 seconds 568 milliseconds to "T" at 3 minutes 01 seconds 122 milliseconds. After the user slides the music playback progress bar or the music spectrogram to 2 minutes 23 seconds or 2 minutes 24 seconds or 3 minutes 01 seconds or 3 minutes 02 seconds, it is difficult to accurately control to continue to slide the music playback progress bar or the music spectrogram to select "O" at 2 minutes 23 seconds 568 milliseconds to "T" at 3 minutes 01 seconds 122 milliseconds.

[0038] This application provides an artificial intelligence assisted music editing method, the process of which is as follows: Figure 2 As shown, the following steps are included: First, based on the basic structural information of the music, a music database is constructed, wherein the music database includes: a music hierarchy structure, i.e., a permutation and combination hierarchy between phrases, sections, and pre-interlude segments, a phrase containing lyrics, a section containing lyrics, and corresponding word-by-word text-timestamp data pairs of lyrics, and one or more pre-interlude segment start and end timestamps; like Figure 3 and Figure 4 As shown, the user selects a piece of music as the selected music and enters the music editing interface. When the user slides the lyrics to start selecting a music clip, the terminal obtains the hierarchical structure of the music in the database, that is, the arrangement and combination hierarchy between the music segment, the phrase, and the pre-interlude clip, the lyrics text-timestamp data pair, one or more pre-interlude clip start and end timestamps, and displays the clip selection control. The user can trigger the pre-interlude selection control to trigger the terminal to quickly select the music clip as the final selected music segment according to the pre-interlude music data.

[0039] Specifically, when the selected fragment selected by the user for the first time is multiple phrases, and the starting point of the selected fragment is the same as the starting point of the segment or the end point of the selected fragment is the same as the end point of the segment, the user is prompted whether to select the corresponding segment as the selected fragment. If the user chooses no, the corresponding multiple phrases will be used as the selected fragment; if the user chooses yes, the segment covering the above multiple phrases will be reset as the selected fragment; when the reset selected fragment is a segment, and the starting point of the selected fragment is the same as the end point of one or more pre-interlude segments or the end point of the selected fragment is the same as the starting point of one or more pre-interlude segments, the user is prompted whether to select the pre-interlude selection control. If the user chooses no, the corresponding segment will be used as the target music for output; if the user chooses yes, the segment and pre-interlude segment will be reset as the target music for output.

[0040] like Figure 3 and Figure 4 As shown in FIG. 1 , a user expects to select a certain segment of a certain song. For example, the user expects to select a certain segment of Jay Chou's "Qilixiang". Figure 5As shown, the song is sequentially connected by (a) prelude, (b) verse paragraph 1, (c) verse paragraph 2, (d) chorus paragraph 1, (e) chorus paragraph 2, (f) interlude, (g) chorus paragraph 1, (h) verse paragraph 2, (i) chorus paragraph 1, (j) chorus paragraph 2, (k) coda.

[0041] A When the segment expected to be selected by the user is two musical phrases in (h), i.e., "O P... Q, R... S T", i.e., "that" at 2 minutes 53 seconds 568 milliseconds to "T" at 3 minutes 01 second 122 milliseconds. The user slides the music lyrics graph to any text in "O P... Q" and any text in "R... S T" of this segment, triggering the terminal to automatically select the music segment from 2 minutes 53 seconds 568 milliseconds to 3 minutes 01 second 122 milliseconds, i.e., "ABCD... EFG,...,..., R... S T", on the playback progress bar or music spectrogram.

[0042] B When the segment expected to be selected by the user is multiple musical phrases in (g) and (h), i.e., A B C D... E F G,...,..., X... Y Z, i.e., "A" at 1 minute 22 seconds 362 milliseconds to "T" at 3 minutes 01 second 122 milliseconds. The user slides the music lyrics graph to any text in "A B C D... E F G" and any text in "X... Y Z" of this segment, triggering the terminal to automatically select the music segment from 1 minute 22 seconds 362 milliseconds to 3 minutes 01 second 122 milliseconds, i.e., "A B C D... E F G,...,..., X... Y Z", on the playback progress bar or music spectrogram.

[0043] At this time, "X... Y Z" of the selected music segment in this example is the same as the starting point of musical phrase (h). It is prompted on the music editing page whether to select the corresponding musical phrases (g) and (h) as the selected segment. If the user selects no, the corresponding multiple musical phrases are taken as the selected segment; if the user selects yes, the musical phrase that overrides the above multiple musical phrases is reset as the selected segment.

[0044] At this time, if the user selects yes and takes the musical phrases (g) and (h) as the selected segment, there is an interlude segment (f) on the left side of (g) in this example. The pre-interlude selection control is displayed on the music editing interface (a pop-up window can also be displayed to prompt the user to select "yes" or "no" to use the pre-interlude selection control. If the user selects "yes"), and the user clicks the pre-interlude selection control ( Figure 4), triggering the terminal to automatically select the music clip containing the music section and the interlude from 1 minute 03 seconds 129 milliseconds to 3 minutes 01 seconds 122 milliseconds on the playback progress bar or the music spectrogram as the output target music, namely (f) (g) (h).

[0045] By using the artificial intelligence-assisted music editing method provided in this application, users can accurately select the desired music clips, which is convenient and quick.

[0046] This application also provides a further artificial intelligence assisted music editing method, such as Figure 6 As shown, the music selection interface displays: music clip control; the music clip selection interface displays: music clip selection control. After the user selects the target music, the music clip control is triggered to enter the music clip interface. In the music clip interface, the user can select the music clip by controlling the music clip selection control. For example, drag the lyrics scroll bar up and down to clip the music.

[0047] Users can also Figure 6 As shown, after selecting the target music, it directly jumps to the music editing interface. The music selection interface displays: music selection control; the music editing interface displays: music clip selection control. The user selects the desired target music, clicks to trigger the music selection control to enter the music editing interface of the target music, and in the music editing interface, the user can confirm the selected clip by controlling the music clip selection control. For example, drag the lyrics scroll bar up and down to clip the music.

[0048] In the related art, the selection of music clips is mainly achieved by controlling the music clip selection control, but due to the limited length of the music clip selection control, it is difficult to accurately locate the music clip the user wants through sliding operation, and the user cannot accurately select the target music clip.

[0049] The artificial intelligence-assisted music editing method provided in the present application adds a pre-interlude selection control to the music editing interface based on the pre-interlude data already stored in the server. The user can quickly and accurately select the target music clip by triggering the pre-interlude selection control.

[0050] Music editing mainly involves the selection, deletion, integration and modification of musical materials. Specifically, this may cover the selection of phrases, the extraction of preludes and the change of sequence. The editing process can also achieve innovation through the superposition of two or more music clips. The editing result can be either a certain degree of adaptation of the original music or a completely new mashup creation. For example, music editing can form new music clips by selecting specific phrases or disconnecting the music from a specific position; then, these new music clips can be rearranged with the original music clips, or even repeated; finally, two different music clips can be superimposed to create new musical effects.

[0051] The music editing interface usually appears on mobile phones, computers, tablets and other devices to implement the music editing function. It can exist as an independent interface or as a part of other interfaces for music editing. For example, Figure 7 It shows an independent music editing interface, and Figure 1 An interface containing a music editing area is displayed, which corresponds to the music editing interface.

[0052] The music editing interface is mainly composed of the target music display control and the segment selection control. The target music display control is the component that controls the display of the target music content, and the segment selection control is the component that controls the selection of phrases and prelude segments.

[0053] The target music refers to the complete music that has not been edited in the current music editing interface. For example, the target music is the music that has been selected before the user enters the music editing interface.

[0054] Controls are any visible visual elements or components on the application user interface, such as pictures, input boxes, text boxes, buttons, labels, etc. Some of them are used to display information, such as picture controls that can display the singer's photo and album cover of a musical work, and text box controls that can display the bibliographic items and lyrics of a musical work; some are used to respond to user actions, such as sliding a clip selection control to select the corresponding music clip, and clicking a play status switch control to change the play status of a musical work.

[0055] The music display control is designed to display the multiple elements of music and has playback control functions (including progress display and playback status adjustment).

[0056] The segment selection control is mainly responsible for the selection of music segments. This control can receive various selection operations from the user, such as single click, double click, slide, drag, long press, and voice commands. In the music editing interface, the music display control and the segment selection control are usually integrated into a single component to facilitate user operation. For example, Figure 7 As shown in the figure, the segment selection control is combined with the music display control, and the user can easily select the desired music segment by sliding the segment selection control. At the same time, when the user slides the segment selection control, the progress of the music selected on the music display control will also change accordingly, thereby intuitively displaying the music segment selected by the user.

[0057] Reference Figure 7 For example, the terminal first presents the initial state of the music editing interface, which mainly includes two components: a music display control and a clip selection control.

[0058] Step 704: When the user performs a selection operation on the segment selection control, the system will obtain the selected music segment accordingly.

[0059] The terminal captures the selection operation from the segment selection control and obtains the corresponding music segment based on this operation.

[0060] The selection operation can be performed in various forms, such as single click, double click, slide, drag, long press, and voice command. Specifically, single click and double click can select a music clip by clicking the lyrics text or its start and end positions once or twice; slide can select a music clip by sliding the lyrics graph to determine the start and end positions of the lyrics text; and voice command can select the corresponding music clip by identifying the lyrics text in the voice command. For example, Figure 7 As shown, the selected operation is the sliding operation on the fragment selection control.

[0061] The method for the system to obtain the selected music clip is as follows: after receiving the clip selection operation, a focus display area will appear on the music display control to display the selected music clip. The focus display method may include: displaying a selection box on the music display control, and highlighting the part of the music lyrics image inside the selection box as the selected music clip. The selection box is usually composed of one or more rectangles; providing conspicuous distinguishing color processing, highlighting and selection mark functions on the music lyrics image; marking the starting position of the selected music clip on the music lyrics image, etc. For example, Figure 7 As shown, a selection box is provided on the syllable editing interface, and the music lyrics image in the box represents the selected music segment.

[0062] The method for obtaining the selected music segment based on the selection behavior may include: after receiving the segment selection operation, immediately identifying and extracting the corresponding music segment. For example, by locating the position of the click or double-click operation, one or more phrases after it are identified as the selected music segment; or by detecting the range of the sliding operation, one or more phrases covered by it are identified as the selected music segment; or by parsing the voice instruction, the specified music segment is identified as one of the selected music segments.

[0063] The clipped music clip is a part or the whole music extracted by the user from the target music. This part or the whole music may be a music clip, such as a phrase, several phrases, or even one or several phrases played in a loop. For example, in a short video production application, the selected music clip may be selected by the user as the background music of the short video, or it may be the music that the user intends to further process.

[0064] For example, Figure 7As shown, in the first state of the music editing interface, when the terminal receives a selection operation (sliding operation) on the segment selection control, it jumps to the second state of the music editing interface, and obtains the selected music segment according to the selection operation. The selected music segment is "ABC D.......X......YZ" in the target music, that is, the music segment corresponding to "A" at 1 minute 22 seconds 362 milliseconds to "Z" at 3 minutes 01 seconds 120 milliseconds.

[0065] The pre-fill selection controls showing the selected music clip of the target music.

[0066] The terminal displays a prelude selection control for the selected music segment of the target music.

[0067] The pre-interlude segment is a pre-interlude segment in the target music identified by the pre-trained neural network. There can be one or more pre-interlude segments.

[0068] The prelude selection control is a control for resetting the selected music segment according to the prelude music segment. The prelude selection control can be a picture, pattern, text box, button, label, etc. displayed on the user interface; it can also be a part of the user interface, which can identify the trigger operation and has the function of resetting the selected music segment according to the prelude music segment. For example, Figure 7 As shown, there is a prelude selection control. Exemplarily, there can be one or more prelude selection controls.

[0069] For example, Figure 7 As shown, in the second state of the music editing interface, a prelude selection control of the prelude segment of the target music is displayed.

[0070] When a trigger operation on the pre-interlude selection control is received, the selected music segment is reset according to the pre-interlude segment.

[0071] The terminal receives a trigger operation on the pre-interlude selection control and resets the selected music segment according to the pre-interlude music segment.

[0072] Resetting the selected music clip according to the pre-interlude music clip means: reselecting the starting point (and end point) of the selected music clip according to the starting point (and end point) of the pre-interlude music clip. For example, reselecting the starting point of the selected music clip according to the starting point of the pre-interlude music clip, and reselecting the end point of the selected music clip according to the end point of the pre-interlude music clip.

[0073] In some possible embodiments, for example, other music clips in the target music determined by using the pre-interlude music clip are determined as the selected music clips, and the music clip between two pre-interlude music clips can be determined as the selected music clip, or the music clip in the target music that starts at a certain time point before the pre-interlude music clip is determined as the selected music clip, or the music clip in the target music that starts at a certain time point after the pre-interlude music clip is determined as the selected music clip.

[0074] For example, refer to Figure 7 In the second stage of the music editing interface, after the terminal receives the trigger command of the prelude selection control, it immediately switches to the third stage of the interface. At this time, according to the pre-set rules, the selected music clip will be updated to the content between 1 minute 03 seconds and 3 minutes 01 seconds in the target music.

[0075] In summary, this implementation scheme introduces more detailed music structure information and pre-interlude selection controls, so that users can more easily and quickly use the preset music structure information when selecting target music clips, and according to the triggering operation of the pre-interlude selection control, the terminal automatically completes the rapid selection of music clips. This implementation scheme greatly reduces the difficulty of users manually locating music clips, making music editing work more convenient and efficient.

[0076] like Figure 8 As shown, this technical solution shows another implementation process of artificial intelligence-assisted music editing operation, which mainly consists of the following four parts: In the step setting stage, a clear and concise music editing page is first presented, which includes a target music display control and a segment selection control, so as to realize the functions of target music display control and segment selection control; The user operation link receives and analyzes the operation instruction of the segment selection control to obtain the selected music segment; Data collection stage 1: accurately obtain the start and end position timestamps of the selected music clips; Data processing stage 2: integrate and analyze the above data to provide support for subsequent operations.

[0077] Specifically, when the start and end positions selected by the user are within a music segment, the selected segment is a phrase or a combination of phrases; when the start and end positions selected by the user span a music segment, the selected segment is a combination of multiple phrases or multiple music segments or multiple music segments and a prelude. More specifically, the selected segment is obtained by comparing and analyzing the music segment hierarchical structure information stored in the database, that is, the arrangement and combination hierarchy between the music segment, the music phrase, and the prelude, the specific music segment start and end timestamps, the phrase start and end timestamps, and the phrase lyrics word-by-word text timestamps with the start and end text-timestamps of the selected music segment obtained by parsing the user operation instruction of the terminal.

[0078] When the selected fragment selected by the user for the first time is multiple phrases, and the starting point of the selected fragment is the same as the starting point of the segment or the end point of the selected fragment is the same as the end point of the segment, the user is prompted whether to select the corresponding segment as the selected fragment. If the user chooses no, the corresponding multiple phrases will be used as the selected fragment; if the user chooses yes, the segment covering the above multiple phrases will be reset as the selected fragment; when the reset selected fragment is a segment, and the starting point of the selected fragment is the same as the end point of one or more pre-interlude segments or the end point of the selected fragment is the same as the starting point of one or more pre-interlude segments, the user is prompted whether to select the pre-interlude selection control. If the user chooses no, the corresponding segment will be used as the target music for output; if the user chooses yes, the segment and pre-interlude segment will be reset as the target music for output.

[0079] like Figure 1 As shown, when a user wants to select a certain segment of a song: for example, the user wants to select a certain segment of Jay Chou's "Qilixiang", the song is composed of (a) prelude, (b) verse section 1, (c) verse section 2, (d) chorus section 1, (e) chorus section 2, (f) interlude, (g) chorus section 1, (h) verse section 2, (i) chorus section 1, (j) chorus section 2, and (k) epilogue.

[0080] AWhen the user expects to select the two phrases in (h), Right now" OP...Q. R......ST", That is, "O" at 2 minutes 53 seconds 568 milliseconds to "T" at 3 minutes 01 seconds 122 milliseconds. The user slides the music lyrics chart to any text of "OP ...... Q" and any text of "R...... ST" in the segment, triggering the terminal to automatically select the music segment of "OP ...... Q, .. R...... ST" from 2 minutes 53 seconds 568 milliseconds to 3 minutes 01 seconds 122 milliseconds on the play progress bar or the music spectrogram.

[0081] B When the user expects to select two music segments (g) and (h), That is, "ABCD...EFG,...,X...YZ", That is, "A" at 1 minute 23 seconds 568 milliseconds to "Z" at 3 minutes 01 seconds 122 milliseconds. The user slides the music lyrics chart to any text of "ABCD......EFG" and any text of "X......YZ" in the segment, triggering the terminal to automatically select the music segment from 1 minute 23 seconds 568 milliseconds to 3 minutes 01 seconds 122 milliseconds "ABC D......EFG, ......., X......YZ" on the play progress bar or music spectrogram.

[0082] At this time, in this example, there is an interlude segment (f) on the left side of (g), and a pre-interlude selection control is displayed on the music editing interface. The user clicks the pre-interlude selection control, triggering the terminal to automatically select the music segment containing the music segment and the interlude from 1 minute 03 seconds 129 milliseconds to 3 minutes 01 seconds 122 milliseconds on the playback progress bar or the music spectrogram as the output target music, namely (f) (g) (h).

[0083] The starting point of the selected segment is the start timestamp of the selected music segment. The terminal obtains the starting point of the selected segment of the selected music segment by receiving the selection operation on the segment selection control. After receiving the selection operation on the segment selection control, the terminal determines the selected music segment and determines the timestamp when the selected music segment is played sequentially as the starting point of the selected segment. Exemplarily, when the selection operation on the segment selection control is a sliding segment selection control, and the music segment in the selected box is determined as the selected music segment, the starting point of the selected segment is the timestamp of the first selected lyrics text when the music segment in the selected box is played sequentially.

[0084] The start and end points of the pre-interlude segment are the start and end timestamps of the pre-interlude music segment. The terminal obtains the start and end points of the pre-interlude music segment by inputting the music file into the pre-trained neural network to obtain the pre-interlude music segment timestamp data, or by storing the pre-interlude music segment timestamp data obtained by inputting the music file into the pre-trained neural network in the terminal and reading it, or by storing the pre-interlude music segment timestamp data obtained by inputting the music file into the pre-trained neural network in the server and receiving the data sent by the server. The pre-interlude music segment data includes the start and end timestamps of the pre-interlude music segment.

[0085] For example, Figure 7As shown, after the terminal receives the selection operation on the clip selection control, in the second state of the music editing interface, the selected music clip is obtained as the music clip corresponding to "A" at 1 minute 23 seconds 560 milliseconds to "Z" at 3 minutes 01 seconds 120 milliseconds in the target music, and the starting point of the selected clip is obtained as 2 minutes 23 seconds 560 milliseconds of the target music. And the starting and ending points of the pre-interlude music clip are obtained as [……, [1 minute 03 seconds 129 milliseconds, 1 minute 23 seconds 560 milliseconds], [4 minutes 01 seconds 120 milliseconds, 4 minutes 31 seconds 122 milliseconds]……] of the target music.

[0086] When the starting point of the selected segment is the same as the end point of the previous interlude segment or the starting point of the music segment, and the end point of the selected segment is the same as the starting point of the previous interlude segment or the end point of the music segment, the corresponding music segment is selected as the selected segment; then a prompt is given as to whether to select the previous interlude selection control. If the user selects no, the corresponding music segment is used as the target music for output; if the user selects yes, the music segment and the previous interlude segment are reset as the target music for output.

[0087] When the starting point of the selected segment is the same as the starting point of the previous interlude segment and / or the ending point of the selected segment is the same as the ending point of the previous interlude segment, the corresponding music segment and the previous interlude segment are output as the selected segment, and the music segment and the previous interlude segment are reset as the output target music.

[0088] Exemplarily, the determination of whether the starting point of the selected segment is the same as the ending point of the pre-interlude segment and whether the ending point of the selected segment is the same as the starting point of the pre-interlude segment can be completed by the terminal and / or the server. When the determination of whether the starting point of the selected segment is the same as the starting and ending points of the pre-interlude segment and whether the ending point of the selected segment is the same as the starting point of the pre-interlude segment is completed by the server, the server sends the determination result to the terminal.

[0089] When the starting point of the selected segment is the same as the ending point of the pre-interlude segment or the ending point of the selected segment is the same as the starting point of the pre-interlude segment, the terminal displays the pre-interlude selection control of the pre-interlude music segment. Exemplarily, when the pre-interlude selection control is an invisible control, when the starting point of the selected segment is the same as the ending point of the pre-interlude segment or the ending point of the selected segment is the same as the starting point of the pre-interlude segment, the terminal opens the pre-interlude selection control of the pre-interlude music segment so that it can receive a trigger operation.

[0090] For example, Figure 7 As shown, if the starting point of the selected segment is the same as the ending point of the pre-interlude segment or the ending point of the selected segment is the same as the starting point of the pre-interlude segment, a pre-interlude selection control of the pre-interlude music segment is displayed in the second state of the music editing interface.

[0091] When a trigger operation on the pre-interlude selection control is received, the selected music segment is reset according to the pre-interlude music segment.

[0092] The terminal receives a trigger operation on the pre-interlude selection control and resets the selected music segment according to the pre-interlude music segment.

[0093] To sum up, the method provided in this embodiment obtains the timestamp starting point of the selected music clip and the starting and ending points of the pre-interlude music clip. When the starting point of the selected clip and the ending point of the pre-interlude are the same, a pre-interlude selection control is added, so that the pre-interlude selection control is only displayed when there is a pre-interlude before and after the selected music clip, thereby improving the utilization rate of the user interface. The user can quickly select a music clip by triggering the pre-interlude selection control, without having to manually locate the music clip multiple times. Artificial intelligence is used to assist in fast and efficient music editing, thereby improving the flexibility, usability and efficiency of selecting music clips.

[0094] Exemplarily, the terminal obtains the start and end points of the prelude segment from the server.

[0095] Fig. 9 A flowchart of an artificial intelligence-assisted music editing method provided by another exemplary embodiment of the present application is shown. The method comprises the following steps: The music editing interface is displayed, and the music editing interface includes: a music display control and a clip selection control for the target music.

[0096] The terminal displays a music editing interface, which includes: a music display control and a segment selection control for the target music.

[0097] When a selection operation on the segment selection control is received, a selected music segment is obtained according to the selection operation.

[0098] The terminal receives a selection operation on the segment selection control, and obtains a selected music segment according to the selection operation.

[0099] The starting point of the selected music segment is obtained, and the pre-interlude timestamp of the target music is obtained from the server, wherein the pre-interlude timestamp includes the starting and ending points of the pre-interlude music segment corresponding to the target music.

[0100] The terminal obtains the starting point of the selected music segment and obtains the pre-interlude timestamp of the target music from the server. The pre-interlude timestamp includes the starting and ending points of the pre-interlude music segment corresponding to the target music.

[0101] In this application, the pre-interlude timestamp is used to indicate the relevant information of the pre-interlude part of the target music. For example, the timestamp includes the start and end position timestamps of the pre-interlude segment.

[0102] The method for obtaining the prelude and interlude timestamp of the target music is that the terminal sends a request to the server to obtain the prelude and interlude timestamp of the target music, and then the server feeds it back to the terminal.

[0103] If the starting point of the selected segment is the same as the ending point of the pre-interlude segment or the ending point of the selected segment is the same as the starting point of the pre-interlude segment, the pre-interlude selection controls of the pre-interlude music segment will be displayed.

[0104] When the starting point of the selected segment is the same as the ending point of the pre-interlude segment or the ending point of the selected segment is the same as the starting point of the pre-interlude segment, the terminal will display the pre-interlude selection control of the pre-interlude music segment.

[0105] When a trigger operation on the pre-interlude selection control is received, the selected music segment is reset according to the pre-interlude music segment.

[0106] The terminal receives a trigger operation on the prelude selection control and adjusts the selected music segment accordingly.

[0107] In summary, the method proposed in this embodiment obtains the start and end points of the selected music clip and the pre-interlude timestamps of the target music (i.e., the start and end timestamps of the pre-interlude music clip), and when the start point of the selected clip is the same as the end point of the pre-interlude clip or the end point of the selected clip is the same as the start point of the pre-interlude clip, adds a pre-interlude selection control, so that the control is only displayed under the above conditions, thereby improving the utilization efficiency of the user interface. The user can quickly select the music clip by triggering the pre-interlude selection control, avoiding multiple manual positioning, achieving the high efficiency of artificial intelligence-assisted music editing, and further enhancing the practicality and efficiency of music clip selection.

[0108] Fig.10 A flowchart of an artificial intelligence-assisted music editing method provided by another exemplary embodiment of the present application is shown in detail. The method comprises the following steps: The music editing interface is displayed first, including a music player for the target music and controls for selecting a clip.

[0109] After entering the interface using a terminal, the user can select the music clip to be edited by clicking the corresponding control.

[0110] When the user selects the required segment, the system automatically obtains the segment and its start and end positions.

[0111] Afterwards, the start and end positions of the previously appeared music clips are obtained for subsequent processing.

[0112] Afterwards, if the starting position of the new segment coincides with the ending position of the previous segment or the end point of the new segment is at the starting position of the previous segment (that is, the two segments are completely connected), the system will display the prelude and interlude selection controls of the previous segment for editing.

[0113] After that, when the user selects the control again, the system will adjust the starting position of the new segment to the starting position of the previous segment to achieve seamless connection.

[0114] For example, Figure 7 As shown, after the user presses the prelude selection button in the second stage of the music editor interface, the system will automatically switch to the third stage and determine that the starting position of the selected music clip corresponds to the starting point of the prelude clip and the ending position corresponds to the ending point of the prelude clip. Specifically, the selected music clip starts from 2 minutes 03 seconds 129 milliseconds in the target music and lasts to 3 minutes 31 seconds 122 milliseconds. After that, the music editor interface will switch to the third stage.

[0115] For example, when there are at least two alternative pre-interlude music clips, and the starting or ending points of these alternative clips are the same, this function will set the starting part of the selected music clip to the starting and ending positions of the shorter pre-interlude clip.

[0116] To sum up, the method provided in this embodiment, by adding a pre-interlude selection control, when the user triggers the control, evaluates the starting and ending points of the music clip selection, and sets them to the corresponding time periods of the pre-interlude music clip respectively, so as to quickly select the music clip and avoid the tedious manual positioning process, making the application of artificial intelligence in music editing more efficient and convenient, and improving the practicality and efficiency of music clip selection.

[0117] For example, when the music clip selection method provided by the above exemplary embodiment is applied to a video editing program (such as a short video sharing or production program), its music clip selection strategy is applied to the audio part (such as adding a background soundtrack to the video file), and considering the duration of the video itself, the music used as accompaniment must have a fixed duration limit. Therefore, when adopting this music clip selection strategy, two steps are involved to determine the start and end time points of the edited music clip.

[0118] Methods for determining the start and end timestamps of a clipped music segment include: After the user completes the short video recording / uploading operation, he can enter the music editing page. As an example, the music clip selection strategy recommended by this application also includes: Step 1: Extract the video duration of the short video.

[0119] The terminal device obtains the video duration of the short video, which is the playback time of the short video.

[0120] In addition, the strategy for adjusting the selected music segment according to the pre-interlude music segment in step 704 described above may also be as follows: Step 2, based on the video length of the short video, the target music length, the timestamp of the selected music clip and the timestamp of the pre-interlude, calculate the timestamp of the edited music clip and the alignment timestamp of the edited music clip and the short video.

[0121] The terminal device determines the timestamp of the edited music segment according to the video length of the short video, the length of the selected music segment, and the starting and ending points of the prelude and interlude segments.

[0122] Methods for determining the timestamp of a clipped music segment include: The duration of the selected music clip is subtracted from the video length of the short video to obtain the duration after the subtraction. For example, the duration of the selected music clip is 37 seconds 554 milliseconds, and the video length of the short video is 1 minute 30 seconds 779 milliseconds, and the duration after the subtraction is 53 seconds 225 milliseconds.

[0123] When the duration after subtraction is less than or equal to 0, the start point of the edited music segment is set as the start point of the selected music segment, and the end point of the selected music segment is reset to the timestamp obtained by adding the start point of the selected music segment to the video length of the short video. That is, for example, the start and end timestamps of the selected music segment are 2 minutes 23 seconds 568 milliseconds to 3 minutes 01 seconds 122 milliseconds, the duration is 37 seconds 554 milliseconds, and the video length of the short video is 30 seconds 554 milliseconds. The duration after subtraction is (-7) seconds, and the end point of the selected music segment is determined to be 2 minutes 23 seconds 568 milliseconds plus 30 seconds 554 milliseconds, that is, 2 minutes 54 seconds 122 milliseconds of the target music is set as the end point of the edited music segment.

[0124] When the duration after subtraction is greater than 0 and less than a threshold greater than 0, the start point of the clipped music segment is set as the start point of the selected music segment, and the end point of the clipped music segment is reset to the end point of the selected music segment. That is, for example, the start and end timestamps of the selected music are 2 minutes 23 seconds 568 milliseconds to 3 minutes 01 seconds 122 milliseconds, the duration is 37 seconds 554 milliseconds, and the video length of the short video is 38 seconds 854 milliseconds. The duration after subtraction is 1 second 300 milliseconds, and the 3 minutes 01 seconds 122 milliseconds of the selected music is set as the end point of the clipped music segment.

[0125] When the subtracted duration is greater than a threshold greater than 0, the method for determining the start and end timestamps of the clipped music segment is: Traverse the duration of the left and right pre-interludes of the selected music clip, and the added duration of the left and right pre-interludes, and obtain the clip with the closest duration after subtraction as the pre-interlude clip, set the starting point of the edited music clip to the smaller of the starting point of the selected music clip and the starting point of the pre-interlude clip, set the starting point of the edited music clip to the smaller of the end point of the selected music clip and the end point of the pre-interlude clip, and set the end point of the edited music clip to the larger end point of the pre-interlude clip and the starting point of the selected music clip. Exemplarily, the length of the left front interlude of the selected music is 20 seconds 439 milliseconds, the length of the right front interlude is 30 seconds 0 milliseconds, and the length of the front interlude on both sides is 50 seconds 439 milliseconds, while the start and end timestamps of the selected music are 2 minutes 23 seconds 568 milliseconds to 3 minutes 01 seconds 122 milliseconds, with a duration of 37 seconds 554 milliseconds, and the video length of the short video is 58 seconds 854 milliseconds. The subtraction of the short video duration from the selected music duration is 30 seconds 300 milliseconds, and the closest segment is the right front interlude. The starting point of the clipped music segment is set to the starting point of the selected music segment, and the end point of the clipped music segment is reset to the end point of the right front interlude.

[0126] The terminal determines the alignment timestamp between the clipped music segment and the video length of the short video according to the video length of the short video and the time length of the clipped music segment.

[0127] In summary, the method provided in this embodiment provides a complete method flow for applying the music clip selection method provided in this application to practice. By adding a prelude selection control, when a user selects a music clip, the user can quickly select the music clip by triggering the prelude selection control, without having to manually locate the music clip multiple times, making artificial intelligence-assisted music editing fast and efficient, and improving the usability and efficiency of selecting music clips.

[0128] The present application also provides a terminal, which includes: a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the artificial intelligence assisted music editing method provided by the above-mentioned method embodiments.

[0129] The present application also provides a computer device, which includes: a processor and a memory, wherein the storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the artificial intelligence assisted music editing method provided by the above-mentioned method embodiments.

[0130] The present application also provides a computer-readable storage medium, which stores at least one instruction, at least one program, code set or instruction set. The at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement the artificial intelligence assisted music editing method provided by the above-mentioned method embodiments.

[0131] It should be understood that the "plurality" mentioned in this article refers to two or more. "And / or" describes the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects are in an "or" relationship.

[0132] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0133] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for artificial intelligence-assisted music editing, characterized in that: The following steps are involved: Based on the basic structural information of the music, a music database is constructed, wherein the music database includes: a music hierarchy structure, a phrase and a segment containing lyrics, a corresponding lyrics text-timestamp data pair, and one or more start and end timestamps of the pre-interlude segment; the music hierarchy structure is an arrangement and combination hierarchy between the segment, the phrase, and the pre-interlude segment; Determine the selection range of the selected music segment, the terminal identifies the lyrics verbatim text-time stamp data and / or pre-interlude music data of the stored phrases and / or music segments in the database corresponding to the selected music segment; and uses the corresponding stored phrases and / or music segments as the selected segment, and the stored pre-interlude segment as the selected pre-interlude segment; Output the target music, the target music contains one or more selected phrases, or, a musical phrase and / or one or more prelude fragments, Or, multiple sections and / or one or more prelude segments.

2. The method for artificial intelligence-assisted music editing according to claim 1, characterized in that: The lyrics verbatim text-timestamp data includes the start and end points of the lyrics text of the music segment accessed from the database and the corresponding time stamp data pair.

3. The method for artificial intelligence-assisted music editing according to claim 1 or 2, characterized in that: The pre-interlude data includes the start and end timestamps of the pre-interlude music segment.

4. The method for artificial intelligence-assisted music editing according to claim 1, characterized in that: The method for determining the selected music clip is: by locating the position of the click or double-click operation, one or more phrases covered by it are identified as the selected music clip; or, by detecting the range of the sliding operation, one or more phrases covered by it are identified as the selected music clip; or, by parsing the voice instruction, the specified music clip therein is identified as one of the selected music clips.

5. The method of artificial intelligence assisted music editing according to claim 4, characterized in that: When the selected segment selected by the user for the first time is multiple phrases, and the starting point of the selected segment is the same as the starting point of the segment or the ending point of the selected segment is the same as the ending point of the segment, a prompt is given to select the corresponding segment as the selected segment. If the user selects no, the corresponding multiple phrases are selected as the selected segments. If the user selects yes, the music segment covering the above-mentioned multiple phrases will be reset as the selected fragment; when the reset selected fragment is a music segment, and the starting point of the selected fragment is the same as the end point of one or more pre-interlude fragments or the end point of the selected fragment is the same as the starting point of one or more pre-interlude fragments, a prompt will be given as to whether to select the pre-interlude selection control. If the user selects no, the corresponding music segment will be used as the target music for output; if the user selects yes, the music segment and pre-interlude fragment will be reset as the target music for output.

6. The method of artificial intelligence assisted music editing according to claim 4, characterized in that: If the starting point of the selected segment is the same as the starting point of the previous interlude segment and / or the ending point of the selected segment is the same as the ending point of the previous interlude segment, the corresponding music segment and the previous interlude segment are output as the selected segment, and the music segment and the previous interlude segment are reset as the output target music.

7. Application of the method of artificial intelligence assisted music editing described in any one of claims 1-6 in video editing.

8. An artificial intelligence-assisted music editing device, characterized in that: include: The display module is used to display a music editing interface, wherein the music editing interface includes: a music display control and a segment selection control of the selected music; and a pre-interlude selection control for displaying the pre-interlude music segment of the selected music; An interaction module, configured to receive a selection operation on the segment selection control; and receive a trigger operation on the prelude and interlude selection control; An acquisition module, configured to, upon receiving a selection operation on the segment selection control, obtain a stored music segment corresponding to the selected music according to the selection operation; The reset module, when receiving a trigger operation on the pre-interlude selection control, resets the selected music segment to obtain the target music according to the selection operation and the pre-interlude segment.

9. A computer device, comprising: A processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement the artificial intelligence-assisted music editing method described in any one of claims 1-6.

10. A computer-readable storage medium, wherein at least one instruction, at least one program, a code set or an instruction set is stored in the storage medium, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the artificial intelligence-assisted music editing method described in any one of claims 1-6.

Images