Queue-based audio playing method and device, equipment and storage medium

Abstract

The invention relates to the technical field of audio equipment control, and discloses a queue-based audio playing method and device, equipment and a storage medium. According to the method, after the audios of the original playlist are randomly disorganized, the played probabilities of the audios are calculated based on the random number generator, and the matched target audios are selected to generate the target playing queue in combination with construction of the user portrait, so that the playlist is associated with the preference of the user; and the audios in the original playlist are dynamically readjusted and sequenced in combination with the playing mode, so that the use habits of the user are better met, the user experience of the audio equipment is improved, and more personalized and intelligent audio services are provided for the user.

Description

Technical Field

[0001] The present application relates to the technical field of audio device control, and in particular to a queue-based audio playback method, apparatus, device and storage medium. Background Art

[0002] With the development of intelligent audio control technology, most smart audio devices and streaming media platforms have realized basic intelligent audio playback, such as providing a playlist function and setting multiple playback modes for the list to realize automatic audio playback control, such as sequential playback, random playback and other modes.

[0003] The random play mode is currently set by R&D personnel and cannot be changed after leaving the factory. The order of the audio list does not change during playback, and the background randomly selects audio to play. This practice does not conform to the usage habits of many users and reduces the user experience. Secondly, the audio list requires users to manually create and manage the playlist, and the system cannot automatically generate or optimize the list based on the user's listening habits and preferences. Summary of the invention

[0004] The main purpose of this application is to solve the problem that the existing audio playlist cannot be dynamically adjusted according to the playback mode.

[0005] The first aspect of the present application provides a queue-based audio playback method, which is applied to an audio device displaying a user interface, and the method includes: responding to a play operation on a play control on the user interface to determine a corresponding play mode; randomly shuffling each audio in an original playlist on the audio device, and inputting the shuffled playlist into a random number generator based on a uniform distribution to calculate a probability of being played, and constructing a new playlist based on the probability of being played; obtaining user information of a user who logs in to the user interface, and constructing a user portrait based on the user behavior habits in the user information; using a content recommendation algorithm to extract target audio that matches the user portrait from the new playlist; sorting all target audio in order of probability of being played from large to small, generating a target playback queue, and playing them.

[0006] Optionally, the audio in the original playlist on the audio device is randomly shuffled, and the shuffled playlist is input into a random number generator based on uniform distribution to calculate the probability of being played, and a new playlist is constructed based on the probability of being played, including: constructing a pseudo-random number generator using a random algorithm; inputting each audio in the original playlist on the audio device into the pseudo-random number generator in sequence to generate a corresponding pseudo-random number, and reordering each audio in the original playlist based on the pseudo-random number to obtain a shuffled playlist; inputting the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played; and randomly selecting two audios from the shuffled playlist according to a preset step size to swap them and generate a new playlist, wherein the difference in the probability of being played of the two randomly selected audios is less than a preset threshold.

[0007] Optionally, the method utilizes a random forward traversal iteration method to randomly select two audios from the shuffled playlist according to a preset step size to swap them and generate a new playlist, including: using each audio in the shuffled playlist as a matrix element, and constructing an audio matrix in order; based on the random forward traversal iteration method, selecting two matrix elements whose playback probability difference is less than a preset threshold from the audio matrix in a one-after-one combination method to swap them until the first matrix element is traversed to generate a new playlist; wherein the one-after-one combination method is that one matrix element is sorted first and one matrix element is sorted last, and the sorting distance between the two matrix elements is equal to the preset step size.

[0008] Optionally, constructing a user portrait based on user behavior habits in the user information includes: extracting the user's playback data from an audio database corresponding to the user interface based on the user information, wherein the playback data includes at least the audio playback history, number of playbacks, collection history, comment history and playback length; utilizing a pre-trained user preference model to extract portrait features from the playback data; and employing a matrix decomposition collaborative filtering algorithm to construct a user portrait based on the portrait features.

[0009] Optionally, the matrix decomposition collaborative filtering algorithm is used to construct a user portrait based on the portrait features, including: clustering the extracted portrait features, and constructing a user-item rating matrix for all portrait features according to the clustering results; factorizing the user-item rating matrix using singular value decomposition to obtain a user factor matrix and a song factor matrix; based on the user factor matrix and the song factor matrix, predicting the score of the user behavior corresponding to the user information, and selecting corresponding portrait features based on the predicted scores to construct the user portrait.

[0010] Optionally, the process of extracting target audio that matches the user portrait from the new playlist using a content recommendation algorithm includes: extracting song features of each audio in the new playlist, wherein the song features include at least song style and song introduction; calculating the similarity between the song style and the song introduction and the user portrait, respectively, and determining the target audio based on the similarity.

[0011] Optionally, the calculating the similarity between the song style and the song introduction and the user portrait respectively, and determining the target audio based on the similarity includes: calculating the cosine similarity between the song style and the song introduction and the user portrait respectively; and selecting N audios with the highest cosine similarities from the audios in the new playlist as the target audios.

[0012] Optionally, all target audios are sorted in order from high to low probability of being played, a target playback queue is generated, and the audios are played, including: all target audios are sorted in order from high to low probability of being played, a target playback queue is generated; playback display effects of the list are extracted based on the user portrait; the target playback queue is synchronized to the user interface, and during playback, the audio being played is marked based on the playback display effects.

[0013] Optionally, after sorting all target audios in order of probability of being played from large to small to generate a target playback queue, the method further includes: constructing an initial population based on the target playback queue, wherein the initial population includes multiple playlist seeds, each playlist seed contains audios of multiple different styles; using a genetic algorithm to select a parent seed from the multiple playlist seeds of the initial population, and iterating based on the parent seed and other playlist seeds to optimize the target playback queue.

[0014] Optionally, the genetic algorithm is used to select a parent seed from multiple playlist seeds of the initial population, and the target play queue is optimized based on the iteration of the parent seed and other playlist seeds, including: calculating the fitness of each playlist seed according to the similarity between each playlist seed and the user portrait and the factors to be optimized; selecting a parent seed from multiple playlist seeds based on the fitness; using a genetic algorithm to cross and mutate the parent seed with other playlist seeds, and selecting a new parent seed based on the results of the crossover and mutation processing, until the similarity of each playlist seed meets the constraint conditions corresponding to the factors to be optimized, and outputting the optimized target play queue.

[0015] Optionally, after sorting all target audios in order from large to small probability of being played, generating a target playback queue, and playing them, it also includes: real-time monitoring of the playback progress of the audio in the target playback queue, and determining whether the playback progress reaches a preset threshold, which is the number of remaining unplayed audios; if so, using a list completion tool to filter target online audios that meet the user portrait from an online audio database, and updating the target online audio to the target playback queue.

[0016] Optionally, the using of a list completion tool to filter target online audio satisfying the user portrait from an online audio database comprises: using the list completion tool to search for online audio from an online audio database using features in the user portrait as an index; extracting song features of each of the online audios, and clustering all of the online audios based on the song features; calculating the similarity between the song features of each category of clustered online audio and the features in the user portrait, and determining the target online audio based on the similarity.

[0017] A second aspect of the present application provides a queue-based audio playback device, which is applied to an audio device displaying a user interface, and the device includes:

[0018] A response module, used to respond to the play operation on the play control on the user interface and determine the corresponding play mode;

[0019] A list generation module, used for randomly shuffling each audio in the original playlist on the audio device, and inputting the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played, and constructing a new playlist based on the probability of being played;

[0020] A portrait generation module, used to obtain information of a user who logs into the user interface and to construct a user portrait based on the user behavior habits in the user information;

[0021] A recommendation module, configured to extract target audio matching the user profile from the new playlist using a content recommendation algorithm;

[0022] The playback module is used to sort all target audios in descending order of playback probability, generate a target playback queue, and play them.

[0023] A third aspect of the present application provides an audio device, comprising a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the audio device executes the queue-based audio playback method as described above.

[0024] A fourth aspect of the present application provides a computer-readable storage medium having instructions stored thereon, and when the instructions are executed by a processor, the queue-based audio playback method as described above is implemented.

[0025] Beneficial effects of this application:

[0026] In the technical solution provided by the present application, the corresponding play mode is determined by responding to the user's play operation on the play control on the user interface; each audio in the original playlist on the audio device is randomly shuffled, and the shuffled playlist is input into a random number generator based on uniform distribution to calculate the probability of being played, and a new playlist is constructed based on the probability of being played; the user information of the user who logs in to the user interface is obtained, and a user portrait is constructed based on the user behavior habits in the user information; the target audio that matches the user portrait is extracted from the new playlist using a content recommendation algorithm; all target audios are sorted in order from large to small in terms of the probability of being played, a target play queue is generated, and the target play queue is played. Such an implementation scheme randomly shuffles the audio in the original playlist, calculates the probability of each audio being played based on a random number generator, and selects the matching target audio to generate a target play queue in combination with the construction of a user portrait, so as to realize the association between the playlist and the user's preference, and dynamically re-arranges the audio in the original playlist in combination with the play mode, which is more in line with the user's usage habits, improves the user experience of the audio device, and provides users with a more personalized and intelligent audio service. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A schematic diagram of a first embodiment of a queue-based audio playback method provided in an embodiment of the present application;

[0028] Figure 2 A schematic diagram of a second embodiment of a queue-based audio playback method provided in an embodiment of the present application;

[0029] Figure 3 A schematic diagram of an original playlist provided in an embodiment of the present application;

[0030] Figure 4 A schematic diagram of a target playback queue provided in an embodiment of the present application;

[0031] Figure 5 A schematic diagram of the structure of a queue-based audio playback device provided in an embodiment of the present application;

[0032] Figure 6 Another structural diagram of a queue-based audio playback device provided in an embodiment of the present application;

[0033] Figure 7 A schematic diagram of the structure of an audio device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to address the defects of the prior art, the present application proposes a queue-based audio playback method, including: in random playback mode, using a high-quality random algorithm to generate and display to the user an actual shuffled playlist rather than the original order; introducing an intelligent list generation algorithm to automatically create a personalized playlist based on the user's historical playback data and preferences, thereby solving the problem that the existing audio playlist cannot be dynamically adjusted according to the playback mode.

[0035] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0036] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 A schematic diagram of a first embodiment of a queue-based audio playback method provided in an embodiment of the present application is applied to an audio device displaying a user interface, where the user interface is an operation interface of an audio playback APP. Based on this, the method specifically includes the following steps:

[0037] 101. In response to a play operation on a play control on a user interface, determine a corresponding play mode.

[0038] In this embodiment, the audio device should be understood as a mobile terminal, such as a mobile phone, a tablet, etc., by real-time monitoring of whether there is a touch operation on the touch screen of the display screen of the audio device. The touch operation can be a click, a long press, etc. When a touch operation is detected, the coordinates of the touch operation are determined, and it is determined whether the coordinates are the position of the playback control for playing audio. If so, it is determined that it is a playback operation for playing audio. The corresponding playback mode is determined based on the triggering of the playback control, and the playback mode includes sequential playback and random playback.

[0039] 102. Randomly shuffle the audio in the original playlist on the audio device, input the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played, and construct a new playlist based on the probability of being played.

[0040] In this embodiment, after determining the playback mode, the original play list of the audio device is obtained in response to the playback operation. The original play list refers to an audio queue created in advance by the user containing multiple audios, and each audio in the audio queue has a certain sorting order. The sorting order can be set manually or automatically sorted according to various time nodes. The time nodes include the collection time of the audio, the release time of the audio, the sorting of the audio name, etc.

[0041] After obtaining the original playlist, all audio in the original playlist are reordered based on the default sorting in the playback mode, and the sorting of each audio in the reordered original playlist is shuffled using a random algorithm. The shuffling here is to use a random algorithm to give each audio a corresponding random number according to the sorting, and sort the audio in the original playlist based on the random number, thereby realizing the random shuffling operation.

[0042] Furthermore, after random shuffling, the shuffled order is input into a random number generator based on uniform distribution. The random number generator performs feature analysis on each audio and generates a random number based on the analyzed features, thereby obtaining the probability of each audio being played, which can also be understood as the probability of the random number being called. Finally, the audio is reordered based on the probability of being played to obtain a new playlist.

[0043] 103. Obtain user information of the login user interface, and build a user profile based on the user behavior habits in the user information.

[0044] In this embodiment, the audio device collects the user account on the user interface and queries the corresponding user information from the audio background. The user information refers to the user's identity information and the playback data of the user's historical audio playback. The playback data includes at least the audio playback history, number of playbacks, collection history, comment history, playback length, etc.

[0045] By analyzing the characteristics of the playback data, a correlation matrix is ​​constructed, and then the correlation matrix is ​​decomposed using matrix decomposition and collaborative filtering to obtain key information such as user preferences. Finally, the corresponding user portrait is constructed based on the decomposed key information. The user portrait should be understood as reflecting the user's behavioral habits towards audio preferences and playback control.

[0046] 104. Use the content recommendation algorithm to extract the target audio that matches the user profile from the new playlist.

[0047] In this embodiment, a content recommendation algorithm is used to extract features of each audio in the new playlist, specifically, the melody, emotion, lyrics keywords and song introduction of each audio are extracted, and then based on the melody, emotion, lyrics keywords and song introduction, the corresponding song portrait is abstracted using the entity abstraction method, and finally the similarity between the song portrait and the user portrait is calculated, so as to filter out the target audio based on the similarity.

[0048] In practical applications, when abstracting a song portrait, the melody and emotion are combined and recognized through a pre-trained song recognition model to determine the music style of the audio. The lyrics keywords and song introduction are recognized through a pre-trained lyrics recognition model to determine the lyrics style of the audio. Finally, the combination of music style and lyrics style is used to determine the specific music type of the audio, and the target audio is determined based on the music type.

[0049] Furthermore, when identifying the style of lyrics, a semantic combination recognition model is also used for identification. That is, through semantic recognition, each lyric keyword is sorted according to similar semantics and related semantics to obtain a vector relationship graph of each lyric keyword. Based on the vector relationship graph, all lyric keywords are sorted and combined to obtain semantics. Based on the semantics and song introduction, they are matched with the existing lyric library to match similar or identical lyrics, thereby obtaining the lyric style.

[0050] 105. Sort all target audios in descending order of probability of being played, generate a target play queue, and play them.

[0051] In this embodiment, after selecting the target audio that satisfies the user portrait, that is, the target audio that satisfies the user preference, the previously calculated probability of being played is obtained to sort each audio to obtain a target playback queue. After obtaining the target playback queue, it replaces the original play list and is displayed on the user interface.

[0052] In another embodiment, after obtaining the target playback queue, the target audio quantity in the target playback queue is compared with the audio quantity in the original playlist. If the two are not equal, the audio in the original playlist is marked based on the target playback queue, that is, the audio in the original playlist that is the same as the target playback queue is marked; then the marked audio is adjusted to the front of the queue according to the order in the target playback queue, and the unmarked audio in the original playlist is arranged after the target playback queue according to the original order.

[0053] Play the audio in order from top to bottom based on the adjusted original playlist.

[0054] Furthermore, while playing audio, it also includes: extracting a playback display effect from a list based on the user portrait; synchronizing the target playback queue to the user interface, and during playback, identifying the audio being played based on the playback display effect.

[0055] This embodiment shuffles the order of the audio in the original playlist after determining the play mode, and then uses a random number generator to estimate the probability of the shuffled audio being played, and selects the target audio in combination with the constructed user portrait, thereby constructing a target play queue. Based on such an audio play scheme, the audio play queue is not the original order, but a real-time order associated with the user's preference, thereby solving the problem that the audio playlist of the prior art cannot be dynamically adjusted according to the play mode, improving the user experience of the smart audio device, and also providing users with more personalized and intelligent audio services.

[0056] See also Figure 2 , a schematic diagram of a second embodiment of a queue-based audio playback method provided in an embodiment of the present application, the method is applied to a mobile phone, an audio playback application is installed on the mobile phone, and the user performs audio playback operations through a user interface on the application, the method specifically includes the following steps:

[0057] 201. In response to a play operation on a play control on a user interface, determine a corresponding play mode.

[0058] The user clicks the play control area on the user interface, and determines the corresponding play mode based on the click operation. If the selected play mode is the random play mode, step 202 is executed. That is, the mobile phone triggers the application to play audio randomly by responding to the user's touch operation on the random play mode button on the user interface.

[0059] 202. Randomly shuffle the audio in the original playlist on the audio device, input the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played, and construct a new playlist based on the probability of being played.

[0060] Specifically, a pseudo-random number generator is constructed using a random algorithm; each audio in an original playlist on an audio device is sequentially input into the pseudo-random number generator to generate a corresponding pseudo-random number, and the audio in the original playlist is reordered based on the pseudo-random number to obtain a scrambled playlist; the scrambled playlist is input into a random number generator based on uniform distribution to calculate the probability of being played; a random forward traversal iteration method is used to randomly select two audios from the scrambled playlist according to a preset step size for swapping, and a new playlist is generated, wherein the difference in the probability of being played of the two randomly selected audios is less than a preset threshold.

[0061] The original playlist is a queue composed of multiple audios that are customized by the user and manually added to match the user's preferences, such as Figure 3 As shown, the random algorithm is then called to shuffle the order of the song audio in the original playlist. The shuffle here is random shuffle, which mainly uses the random algorithm to use the order of each audio as the random number generation condition to generate a pseudo-random number, and then use the pseudo-random number and the audio itself as the input of the random number generator to calculate a random number, and get the probability of the audio being played. A new playlist is constructed based on the probability of being played. If a high-quality random algorithm is used to shuffle the original playlist to generate a new random play sequence, this method uses the std::shuffle algorithm in the modern C++ standard library, which is based on a uniformly distributed random number generator to ensure that the probability of each arrangement is equal.

[0062] In actual applications, when using a random number generator to calculate the probability of each audio being played, the shuffled sorting order of each audio and the audio introduction are extracted as input, and the random number generator calculates the probability of the audio being played using the shuffled sorting order and the audio introduction as random number factors.

[0063] Furthermore, after extracting key information from the audio introduction, a semantic vector is output based on the key information, the semantic vector and the sorting order are input into a random number generator to calculate a random number for the audio, and the random number is used as the probability of the audio being played.

[0064] In another embodiment, the method of using random forward traversal iteration to randomly select two audios from the shuffled playlist according to a preset step size for swapping to generate a new playlist includes: using each audio in the shuffled playlist as a matrix element, and constructing an audio matrix in order; based on the random forward traversal iteration method, selecting two matrix elements whose playback probability difference is less than a preset threshold from the audio matrix in a one-after-one combination manner for swapping until the first matrix element is traversed to generate a new playlist; wherein the one-after-one combination manner is that one matrix element is sorted first and one matrix element is sorted last, and the sorting distance between the two matrix elements is equal to the preset step size.

[0065] For example, when the original playlist of length n is shuffled to construct a new playlist, it starts from i=n-1 and traverses forward to i=1, randomly selecting an index j in the range of [0, i] each time, and then swapping the elements at positions i and j.

[0066] For example, after shuffling the order of the audios in the original playlist and converting them into a matrix [A, B, C, D, E], where A, B, C, D, E represent audios, traverse from E to A, and select one as the exchange object each time you traverse it, and repeat the loop until all audios are traversed.

[0067] Taking the matrix [A,B,C,D,E] as an example, the first iteration: select E and exchange it with C → [A,B,E,D,C]; the second iteration: select D and exchange it with A → [D,B,E,A,C]; the third iteration: select B and exchange it with E → [D,E,B,A,C]; the fourth iteration: select E and exchange it with D → [E,D,B,A,C]; the final new playlist is [E,D,B,A,C].

[0068] Further, after obtaining the new playlist, the generated new playlist is updated to the user interface, so that the user can intuitively see the actual play order of the list in the random play mode. Based on this, the user can continue to operate on the updated playlist, such as skipping songs, returning to the previous song, etc.

[0069] In this embodiment, while the playlist is updated, the relevant information of the original playlist, such as the original order of the audio, can be retained in the background of the application, so that after the user cancels the random play mode, the original playlist can be automatically restored and updated to the user interface.

[0070] 203. Obtain user information of the user who logs into the user interface, and build a user profile based on the user behavior habits in the user information.

[0071] In this embodiment, the user information may be the account information, identity information, etc. of the user when logging into the user interface. Playback data is extracted from the background based on the user information. The playback data includes playback history information and user operation records on the user interface. The user portrait is obtained by analyzing the features in the playback data.

[0072] Specifically, based on the user information, the user's playback data is extracted from the audio database corresponding to the user interface, wherein the playback data at least includes the audio's playback history, number of playbacks, collection history, comment history, and playback length; a pre-trained user preference model is used to extract portrait features from the playback data; and a matrix decomposition collaborative filtering algorithm is used to construct a user portrait based on the portrait features.

[0073] In practical applications, the process of building user portraits can be summarized into the following four steps:

[0074] Step 1: Collect the user's playback data, which includes but is not limited to: playback history, favorite songs, like / dislike records, skip behavior, repeated playback times and playback duration.

[0075] Step 2, data preprocessing, cleans and preprocesses the collected original playback data, which includes operations such as removing outliers and formatting.

[0076] Step 3, feature extraction, extracts user preference features from the preprocessed data, such as the most frequently listened music style, favorite artists, frequently listened time period, and audio duration preference.

[0077] Step 4: Build a user portrait. Based on the extracted features, use a machine learning algorithm to build a user portrait model.

[0078] Optionally, constructing the user portrait is implemented by adopting a matrix decomposition collaborative filtering algorithm, that is, the matrix decomposition collaborative filtering algorithm is used to construct the user portrait based on the portrait features, including: clustering the extracted portrait features, and constructing a user-item rating matrix for all portrait features according to the clustering results; factorizing the user-item rating matrix using a singular value decomposition method to obtain a user factor matrix and a song factor matrix; based on the user factor matrix and the song factor matrix, predicting the score of the user behavior corresponding to the user information, and selecting the corresponding portrait features based on the predicted scores to construct the user portrait.

[0079] In practical applications, all extracted portrait features are clustered to output multiple portrait features of different categories, and then each portrait feature is estimated to obtain the preference level of the song. The preference level is used as a matrix element to construct a user-item rating matrix; based on the user-item rating matrix, singular value decomposition is used to extract the user factor matrix and the song factor matrix.

[0080] Furthermore, before constructing the user portrait, the optimization function is also included to optimize the user factor matrix and the song factor matrix, and the user behavior is scored based on the optimized user factor matrix and the song factor matrix to obtain the user portrait. The optimization function is as follows:

[0081] Σ(r_ui-p_u·q_i^T)^2+λ(||p_u||^2+||q_i||^2)

[0082] Among them, r_ui is the actual score of user u on item i, p_u is the factor vector of user u, q_i is the factor vector of item i, and λ is the regularization parameter.

[0083] Assume there are 3 users and 4 songs, the user-item rating matrix R is:

[0084]

[0085] Through matrix decomposition, we obtain the user factor matrix P and the song factor matrix Q, thereby predicting unknown ratings and constructing user portraits.

[0086] 204. Extract song features of each audio in the new playlist, where the song features at least include song style and song introduction.

[0087] In this step, when extracting song features, a semantic analysis model is used to identify the semantics, emotions, melody and other information of the music and words in the audio, and then the identified information is combined in a directed graph, specifically converting the semantics, emotions, melody and other information into vectors, building a directed graph based on the vectors, and finally matching the graphs to obtain the song style, and the song introduction is searched from the Internet through the audio information. The vector conversion here can be achieved using existing conventional conversion algorithms.

[0088] 205. Calculate the similarity between the song style and the song introduction and the user portrait respectively, and determine the target audio based on the similarity.

[0089] Specifically, the cosine similarities between the song style and the song introduction and the user portrait are calculated respectively; and the N audios with the highest cosine similarities are selected from the audios in the new playlist as the target audios.

[0090] The formula of cosine similarity is: cos(θ)=(A·B) / (||A||||B||), where A and B represent the user profile vector and item feature vector, respectively.

[0091] For example, user profile: [0.8, 0.2, 0.6, 0.4]; song 1 features: [0.9, 0.1, 0.8, 0.3]; song 2 features: [0.2, 0.7, 0.3, 0.8].

[0092] After calculating the similarity between the user portrait and each song, songs with high similarity are selected as target audios recommended to the user.

[0093] 206. Sort all target audios in descending order of probability of being played, generate a target play queue, and play them.

[0094] In this embodiment, all target audios are sorted in descending order of probability of being played, and a target play queue is generated, such as Figure 4 As shown; extracting the playback display effects of the list based on the user portrait; synchronizing the target playback queue to the user interface, and during playback, identifying the audio being played based on the playback display effects.

[0095] In another embodiment, after sorting all target audios in descending order of probability of being played to generate a target play queue, the method further includes:

[0096] An initial population is constructed based on the target playback queue, wherein the initial population includes multiple playlist seeds, each playlist seed contains audio of multiple different styles; a parent seed is selected from the multiple playlist seeds of the initial population using a genetic algorithm, and the target playback queue is optimized based on the iteration of the parent seed and other playlist seeds.

[0097] The method of selecting a parent seed from a plurality of playlist seeds of the initial population by using a genetic algorithm, and optimizing the target play queue based on iteration of the parent seed and other playlist seeds, includes:

[0098] The fitness of each playlist seed is calculated according to the similarity between each playlist seed and the user portrait and the factors to be optimized; a parent seed is selected from multiple playlist seeds based on the fitness; the parent seed is crossover and mutation processed with other playlist seeds using a genetic algorithm, and a new parent seed is selected based on the results of the crossover and mutation processing, until the similarity of each playlist seed meets the constraint conditions corresponding to the factors to be optimized, and the optimized target play queue is output.

[0099] Specifically, during the crossover and mutation processing, a hierarchical elite selection operator is used to select the parent seed of the next iteration, and a crossover algorithm is used to perform seed gene recombination; an iteration termination condition is set, and when the iterative population meets the iteration termination condition, the optimized target playback queue is output.

[0100] For example, the generated recommendation list is optimized by considering at least one of the following factors: the diversity of music styles, the proportion of new content introduced, and recent changes in user preferences.

[0101] The process of using genetic algorithms to optimize the target play queue includes the following steps:

[0102] 1) Initialize the playlist population;

[0103] 2) Evaluate the fitness of each playlist;

[0104] 3) Select excellent individuals as parents;

[0105] 4) Crossover operation: generating offspring from parent generation;

[0106] 5) Mutation operation: randomly changing some genes;

[0107] 6) Repeat steps 2-5 until the termination condition is reached;

[0108] The function for calculating fitness is:

[0109] f(playlist)=Σw_i*score_i(playlist)

[0110] Among them, w_i is the weight of each evaluation indicator, and score_i is the score of the playlist on each indicator.

[0111] Assume the target playback queue is: [1,2,3,4,5], after crossover: [1,2,6,4,5], after mutation: [1,7,6,4,5], through multiple generations of evolution, the optimized target playback queue is finally obtained.

[0112] During the playback of the target playback queue, the system also monitors the user's real-time operations on the audio of the target playback queue, such as skipping songs, playing the next song, etc. The user's preference for the played audio is judged based on the duration of the operation and recorded in the user's playback data for updating the user portrait or adjusting the recommended target audio.

[0113] 207. Monitor the playing progress of the audio in the target playing queue in real time, and determine whether the playing progress reaches a preset threshold, which is the number of remaining unplayed audio.

[0114] By monitoring the progress of the target playback queue in real time, when the playback progress reaches a preset threshold (such as 3 songs remaining), the list completion function is triggered.

[0115] 208. If yes, use a list completion tool to filter target online audio that meets the user profile from an online audio database, and update the target online audio to a target playback queue.

[0116] In this embodiment, a list completion tool is used to search online audio from an online audio database using the features in the user portrait as an index; song features of each online audio are extracted, and all the online audio are clustered based on the song features; the similarity between the song features of each type of clustered online audio and the features in the user portrait is calculated, and the target online audio is determined based on the similarity.

[0117] Specifically, by performing feature analysis on the currently playing target playback queue, key information such as main music style, artists with high frequency of appearance, song emotional characteristics and audio duration distribution are extracted.

[0118] Specifically, the K-means clustering algorithm is used to analyze the characteristics of the queue. The implementation steps are as follows:

[0119] 1) Randomly select K center points;

[0120] 2) Assign each data point to the nearest center point;

[0121] 3) Recalculate the center point of each cluster;

[0122] 4) Repeat steps 2-3 until convergence.

[0123] Assume that the characteristics of the songs in the playlist are:

[0124] Song 1: [0.8, 0.1, 0.1] (popularity, dance music, lyricism)

[0125] Song 2: [0.2, 0.7, 0.1]

[0126] Song 3: [0.5, 0.2, 0.3]

[0127] Through K-means clustering, we can get the main feature clusters, such as [0.7, 0.2, 0.1], which means that this queue is dominated by popular styles.

[0128] Furthermore, after analyzing the features, an intelligent completion strategy is constructed based on the extracted features and user portraits. Among them, the strategy considers factors such as maintaining the consistency of the list style, introducing an appropriate amount of new elements, and being consistent with the user's current context (such as time, location, and activity type).

[0129] According to the completion strategy, qualified candidate audios are retrieved from the audio library, wherein when searching the candidate audios, the above-mentioned cosine similarity can be used to calculate the matching degree between the candidate audios and the target features.

[0130] For example, target features: [0.7, 0.2, 0.1]

[0131] Candidate audio 1: [0.8, 0.1, 0.1], similarity = 0.99

[0132] Candidate audio 2: [0.2, 0.7, 0.1], similarity = 0.45

[0133] Select audio with high similarity as completion candidate and update it after the audio in the target playback queue.

[0134] In this embodiment, during the completion process, in order to ensure the balance and diversity of the overall list, a Markov chain model can be used to achieve this, that is, the probability of each audio in the audio library being transferred to the target playback queue is calculated based on the Markov chain model, and a Markov chain transfer probability matrix is ​​generated: P = [p_ij], where p_ij represents the probability of transferring from state i to state j, and its state transfer equation is: π(t+1) = π(t)*P, where π(t) is the state distribution vector at time t; candidate audio is selected based on probability and updated to the queue.

[0135] Assuming there are three music styles (pop, rock, classical), the transition probability matrix is:

[0136]

[0137] If the current song is a pop song, the probability of choosing a rock song next is 0.2, and the probability of choosing a classical song is 0.1. In this way, the system can introduce a certain amount of diversity while maintaining the consistency of style.

[0138] Furthermore, after the completion is updated, the completion queue is displayed to the user, allowing the user to make adjustments and provide feedback. The system records this feedback and uses it to continuously optimize the completion algorithm.

[0139] This application uses a high-quality std::shuffle algorithm to generate a truly random play sequence, and intuitively displays it on the user interface, which conforms to the user's usage habits and improves the operating experience. Secondly, the method introduces an intelligent list generation function, uses a collaborative filtering algorithm based on matrix decomposition to build a user portrait, combines a content-based recommendation algorithm and a genetic algorithm to optimize, and creates a personalized playlist for the user, significantly improving the accuracy of the recommendation and user satisfaction. Finally, the present invention also develops an innovative list completion function, analyzes the current list features through K-means clustering, uses cosine similarity to retrieve candidate audio, and introduces a Markov chain model to dynamically adjust the completion content, ensuring the coherence and diversity of the playlist. The organic combination of these three functions not only optimizes the user interaction experience, but also realizes intelligent and personalized audio playback services, significantly improving the competitiveness of the product. At the same time, the modular design of the method makes it easy to implement and expand on different intelligent audio devices, and has a wide range of application prospects. In general, the present invention comprehensively improves the user experience of intelligent audio devices through advanced algorithms and intelligent designs, and provides users with more personalized and intelligent audio services.

[0140] The above describes the queue-based audio playback method in the embodiment of the present application. The following describes in detail the queue-based audio playback device in the embodiment of the present application from the perspective of modular functional entities. Figure 5, a schematic diagram of a structure of a queue-based audio playback device provided in an embodiment of the present application, which is applied to an audio device displaying a user interface, and the device includes:

[0141] A response module 510, configured to respond to a play operation on a play control on the user interface and determine a corresponding play mode;

[0142] A list generation module 520 is used to randomly shuffle the audio in the original playlist on the audio device, input the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played, and construct a new playlist based on the probability of being played;

[0143] A portrait generation module 530 is used to obtain information of a user who logs into the user interface and to construct a user portrait based on user behavior habits in the user information;

[0144] A recommendation module 540 is used to extract target audio matching the user portrait from the new playlist using a content recommendation algorithm;

[0145] The playing module 550 is used to sort all target audios in order of probability of being played from large to small, generate a target playing queue, and play them.

[0146] This solution randomly shuffles the audio in the original playlist, calculates the probability of each audio being played based on a random number generator, and selects matching target audio to generate a target playback queue based on the construction of a user profile, thereby associating the playlist with the user's preferences. It also dynamically reorders the audio in the original playlist based on the playback mode, which is more in line with user habits, improves the user experience of audio devices, and provides users with more personalized and intelligent audio services.

[0147] See also Figure 6 , another structural diagram of a queue-based audio playback device provided in an embodiment of the present application, applied to an optical sensor communication base station, the device includes:

[0148] A response module 510, configured to respond to a play operation on a play control on the user interface and determine a corresponding play mode;

[0149] A list generation module 520 is used to randomly shuffle the audio in the original playlist on the audio device, input the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played, and construct a new playlist based on the probability of being played;

[0150] A portrait generation module 530 is used to obtain information of a user who logs into the user interface and to construct a user portrait based on user behavior habits in the user information;

[0151] A recommendation module 540 is used to extract target audio matching the user portrait from the new playlist using a content recommendation algorithm;

[0152] The playing module 550 is used to sort all target audios in order of probability of being played from large to small, generate a target playing queue, and play them.

[0153] In this embodiment, the list generation module 520 includes:

[0154] A first construction unit 521 is used to construct a pseudo-random number generator using a random algorithm;

[0155] A scrambling unit 522 is used to input the audio in the original playlist on the audio device into the pseudo-random number generator in sequence to generate corresponding pseudo-random numbers, and reorder the audio in the original playlist based on the pseudo-random numbers to obtain a scrambled playlist;

[0156] A first calculation unit 523 is used to input the shuffled play list into a random number generator based on uniform distribution to calculate the probability of being played;

[0157] The list generation unit 524 is used to randomly select two audios from the shuffled playlist to swap according to a preset step size using a random forward traversal iteration method to generate a new playlist, wherein the difference in the probability of playing the two randomly selected audios is less than a preset threshold.

[0158] In this embodiment, the list generating unit 524 is specifically used for:

[0159] Using each audio in the shuffled playlist as a matrix element, an audio matrix is ​​constructed in order;

[0160] Based on a random forward traversal iteration method, two matrix elements whose playback probability difference is less than a preset threshold are selected from the audio matrix in a one-after-one combination manner for exchange until the first matrix element is traversed to generate a new playlist; wherein the one-after-one combination manner is that one matrix element is sorted first and one matrix element is sorted last, and the sorting distance between the two matrix elements is equal to the preset step size.

[0161] In this embodiment, the portrait generation module 530 includes:

[0162] A first extraction unit 531 is used to extract the user's playback data from the audio database corresponding to the user interface based on the user information, wherein the playback data at least includes the audio playback history, playback times, collection history, comment history and playback length;

[0163] A second extraction unit 532, configured to extract the portrait features in the playback data using a pre-trained user preference model;

[0164] The second construction unit 533 is used to construct a user portrait based on the portrait features by using a matrix decomposition collaborative filtering algorithm.

[0165] In this embodiment, the second constructing unit 533 is specifically used for:

[0166] Cluster the extracted image features, and construct a user-item rating matrix based on the clustering results of all image features;

[0167] Using singular value decomposition method to factorize the user-item rating matrix to obtain a user factor matrix and a song factor matrix;

[0168] Based on the user factor matrix and the song factor matrix, a score prediction is performed on the user behavior corresponding to the user information, and corresponding portrait features are selected based on the predicted score to construct a user portrait.

[0169] In this embodiment, the recommendation module 540 includes:

[0170] The third extraction unit 541 is used to extract song features of each audio in the new playlist, wherein the song features at least include song style and song introduction;

[0171] The second calculation unit 542 is used to calculate the similarity between the song style and the song introduction and the user portrait respectively, and determine the target audio based on the similarity.

[0172] In this embodiment, the third extraction unit 541 is specifically used for:

[0173] Calculate the cosine similarity between the song style and the song introduction and the user portrait respectively;

[0174] Select N audios with the highest cosine similarity from the audios in the new playlist as target audios.

[0175] In this embodiment, the playing module 550 includes:

[0176] The sorting unit 551 is used to sort all target audios in descending order of probability of being played, and generate a target play queue;

[0177] A fourth extraction unit 552, configured to extract a play display effect of the list based on the user portrait;

[0178] The playback unit 553 is used to synchronize the target playback queue to the user interface and, during playback, identify the audio being played based on the playback display special effects.

[0179] In this embodiment, the audio playback device further includes: an optimization module 560, which is used to:

[0180] Building an initial population based on the target play queue, wherein the initial population includes a plurality of playlist seeds, each playlist seed containing audios of a plurality of different styles;

[0181] A genetic algorithm is used to select a parent seed from a plurality of playlist seeds of the initial population, and the target play queue is optimized based on iteration of the parent seed and other playlist seeds.

[0182] In this embodiment, the optimization module 560 includes:

[0183] The third calculation unit 561 is used to calculate the fitness of each playlist seed according to the similarity between each playlist seed and the user portrait and the factors to be optimized;

[0184] A voting unit 562, configured to vote a parent seed from a plurality of playlist seeds based on the fitness;

[0185] The optimization unit 563 is used to use a genetic algorithm to cross and mutate the parent seed with other playlist seeds, and select a new parent seed based on the results of the crossover and mutation processing, until the similarity of each playlist seed meets the constraint conditions corresponding to the factor to be optimized, and output the optimized target play queue.

[0186] In this embodiment, the audio playback device further includes a completion module 570, which is used to:

[0187] Monitor the playing progress of the audio in the target playing queue in real time, and determine whether the playing progress reaches a preset threshold, which is the number of remaining unplayed audio;

[0188] If so, a list completion tool is used to filter target online audio that meets the user portrait from an online audio database, and the target online audio is updated to the target playback queue.

[0189] In this embodiment, the completion module 570 includes:

[0190] A search unit 571, configured to use a list completion tool to search for online audio from an online audio database using features in the user portrait as indexes;

[0191] A clustering unit 572, configured to extract song features of each of the online audios, and cluster all the online audios based on the song features;

[0192] The determination unit 573 is used to calculate the similarity between the song features of each type of online audio after clustering and the features in the user portrait, and determine the target online audio based on the similarity.

[0193] The embodiment of the present application shuffles the audio order in the original playlist after determining the play mode, and then uses a random number generator to estimate the probability of the shuffled audio being played, and selects the target audio in combination with the constructed user portrait, thereby constructing a target play queue. Based on such an audio playback scheme, the audio play queue is not the original order, but a real-time order associated with the user's preference, thereby solving the problem that the audio playlist of the prior art cannot be dynamically adjusted according to the play mode, improving the user experience of the smart audio device, and also providing users with a more personalized and intelligent audio service.

[0194] above Figure 5 and 6 The queue-based audio playback device in the embodiment of the present application is described in detail from the perspective of modular functional entities, and the audio device in the embodiment of the present application is described in detail from the perspective of hardware processing.

[0195] See also Figure 7 As shown, the audio device includes a processor 700 and a memory 701 . The memory 701 stores machine executable instructions that can be executed by the processor 700 . The processor 700 executes the machine executable instructions to implement the above-mentioned queue-based audio playback method.

[0196] Further, Figure 7 The audio device shown further includes a bus 702 and a communication interface 703 , and the processor 700 , the communication interface 703 and the memory 701 are connected via the bus 702 .

[0197] The memory 701 may include a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), for example, at least one disk storage. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 703 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used. The bus 702 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0198] The processor 700 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 700. The above processor 700 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The disclosed methods, steps and logic block diagrams in the embodiments of the present disclosure can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present disclosure can be directly embodied as a hardware decoding processor to be executed, or a combination of hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 701, and the processor 700 reads the information in the memory 701, and combines its hardware to complete the steps of the queue-based audio playback method provided in the above-mentioned embodiment.

[0199] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the various steps of the queue-based audio playback method provided in the above-mentioned embodiments.

[0200] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described equipment, devices, and units can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0201] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0202] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. 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 application.

Claims

1. A queue-based audio playback method, applied to an audio device displaying a user interface, characterized in that: The method comprises: In response to a play operation on a play control on the user interface, determining a corresponding play mode; Randomly shuffle the audio in the original playlist on the audio device, input the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played, and construct a new playlist based on the probability of being played; Obtaining user information of a user who logs into the user interface, and constructing a user profile based on the user behavior habits in the user information; Extracting target audio matching the user profile from the new playlist using a content recommendation algorithm; All target audios are sorted in descending order of probability of being played, a target play queue is generated, and the target audios are played.

2. The queue-based audio playback method according to claim 1, characterized in that: The step of randomly shuffling each audio in the original playlist on the audio device, inputting the shuffled playlist into a random number generator based on uniform distribution to calculate a probability of being played, and constructing a new playlist based on the probability of being played, comprises: Use random algorithms to build pseudo-random number generators; Inputting the audios in the original playlist on the audio device into the pseudo-random number generator in sequence to generate corresponding pseudo-random numbers, and reordering the audios in the original playlist based on the pseudo-random numbers to obtain a shuffled playlist; Input the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played; By using a random forward traversal iteration method, two audios are randomly selected from the shuffled playlist according to a preset step size for swapping, and a new playlist is generated, in which the difference in the probability of being played of the two randomly selected audios is less than a preset threshold.

3. The queue-based audio playback method according to claim 2, characterized in that: The method of using random forward traversal iteration to randomly select two audios from the shuffled playlist according to a preset step length for swapping to generate a new playlist includes: Using each audio in the shuffled playlist as a matrix element, an audio matrix is ​​constructed in order; Based on a random forward traversal iteration method, two matrix elements whose playback probability difference is less than a preset threshold are selected from the audio matrix in a one-after-one combination manner for exchange until the first matrix element is traversed to generate a new playlist; wherein the one-after-one combination manner is that one matrix element is sorted first and one matrix element is sorted last, and the sorting distance between the two matrix elements is equal to the preset step size.

4. The queue-based audio playback method according to claim 1, characterized in that: The constructing of a user profile based on the user behavior habits in the user information includes: Extracting the user's playback data from the audio database corresponding to the user interface based on the user information, wherein the playback data at least includes the audio's playback history, playback times, collection history, comment history, and playback length; Using a pre-trained user preference model, extracting portrait features from the playback data; A matrix decomposition collaborative filtering algorithm is used to construct a user profile based on the profile features.

5. The queue-based audio playback method according to claim 4, characterized in that: The method of using a matrix decomposition collaborative filtering algorithm to construct a user portrait based on the portrait features includes: Cluster the extracted image features, and construct a user-item rating matrix based on the clustering results of all image features; Using singular value decomposition method to factorize the user-item rating matrix to obtain a user factor matrix and a song factor matrix; Based on the user factor matrix and the song factor matrix, a score prediction is performed on the user behavior corresponding to the user information, and corresponding portrait features are selected based on the predicted score to construct a user portrait.

6. The queue-based audio playback method according to claim 1, characterized in that: The extracting target audio matching the user portrait from the new playlist by using a content recommendation algorithm includes: Extracting song features of each audio in the new playlist, wherein the song features at least include song style and song introduction; The similarities between the song style and the song introduction and the user portrait are calculated, and the target audio is determined based on the similarities.

7. The queue-based audio playback method according to claim 6, characterized in that: The calculating the similarity between the song style and the song introduction and the user portrait respectively, and determining the target audio based on the similarity, includes: Calculate the cosine similarity between the song style and the song introduction and the user portrait respectively; Select N audios with the highest cosine similarity from the audios in the new playlist as target audios.

8. The queue-based audio playback method according to claim 1, characterized in that: The step of sorting all target audios in descending order of probability of being played, generating a target play queue, and playing the target audios comprises: Sort all target audios in descending order of probability of being played, and generate a target play queue; Extracting the play display effects of the list based on the user portrait; The target play queue is synchronized to the user interface, and during playback, the audio being played is identified based on the playback display special effects.

9. The queue-based audio playback method according to any one of claims 1 to 8, characterized in that: After all target audios are sorted in descending order of the probability of being played to generate a target play queue, the method further includes: Building an initial population based on the target play queue, wherein the initial population includes a plurality of playlist seeds, each playlist seed containing audios of a plurality of different styles; A genetic algorithm is used to select a parent seed from a plurality of playlist seeds of the initial population, and the target play queue is optimized based on iteration of the parent seed and other playlist seeds.

10. The queue-based audio playback method according to claim 9, characterized in that: The method of selecting a parent seed from a plurality of playlist seeds of the initial population by using a genetic algorithm, and optimizing the target play queue based on iteration of the parent seed and other playlist seeds, comprises: Calculate the fitness of each playlist seed based on the similarity between each playlist seed and the user portrait and the factors to be optimized; Selecting a parent seed from a plurality of playlist seeds based on the fitness; The parent seed is cross-linked and mutated with other playlist seeds using a genetic algorithm, and a new parent seed is selected based on the results of the cross-link and mutation process until the similarity of each playlist seed satisfies the constraint conditions corresponding to the factors to be optimized, and the optimized target play queue is output.

11. The queue-based audio playback method according to any one of claims 1 to 8, characterized in that: After all target audios are sorted in descending order of probability of being played, a target play queue is generated, and the target audios are played, the method further includes: Monitor the playing progress of the audio in the target playing queue in real time, and determine whether the playing progress reaches a preset threshold, which is the number of remaining unplayed audio; If so, a list completion tool is used to filter target online audio that meets the user portrait from an online audio database, and the target online audio is updated to the target playback queue.

12. The queue-based audio playback method according to claim 11, characterized in that: The using the list completion tool to select target online audio that meets the user profile from the online audio database includes: Using a list completion tool, searching for online audio from an online audio database using features in the user portrait as an index; Extracting song features of each of the online audios, and clustering all the online audios based on the song features; The similarity between the song features of each type of online audio after clustering and the features in the user portrait is calculated, and the target online audio is determined based on the similarity.

13. A queue-based audio playback device, applied to an audio device displaying a user interface, characterized in that: The device comprises: A response module, used to respond to the play operation on the play control on the user interface and determine the corresponding play mode; A list generation module, used for randomly shuffling each audio in the original playlist on the audio device, and inputting the shuffled playlist into a random number generator based on uniform distribution to calculate the probability of being played, and constructing a new playlist based on the probability of being played; A portrait generation module, used to obtain information of a user who logs into the user interface and to construct a user portrait based on the user behavior habits in the user information; A recommendation module, configured to extract target audio matching the user profile from the new playlist using a content recommendation algorithm; The playback module is used to sort all target audios in descending order of playback probability, generate a target playback queue, and play them.

14. An audio device, characterized in that: The audio device includes a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the audio device executes the queue-based audio playback method as described in any one of claims 1-12.

15. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the queue-based audio playback method as described in any one of claims 1-12 is implemented.

Images