Audio novel playing method and system based on cloud space, and Bluetooth earphone

Abstract

The invention discloses an audio novel playing method based on a cloud space, and the method comprises the steps: playing a current audio novel containing an audio advertisement through a first application program, and obtaining the book name and chapter information of the corresponding audio novel; caching follow-up chapter audio of the current chapter according to the book name and the chapter information, removing audio advertisements and uploading the follow-up chapter audio to the cloud space; and when chapter ending feature words appear in the voiced novel, playing subsequent chapter audios in the cloud space through the second application program. The time for the user to listen to the audio advertisement is reduced, and the user experience is improved. And meanwhile, new database contents are continuously obtained through the system, so that the audio advertisement can be more and more efficiently and accurately identified.

Description

Technical Field

[0001] The present invention relates to the technical field of audio advertisement playing, and in particular to a cloud-based audio novel playing method, system and Bluetooth headset. Background Art

[0002] Audio novels can reduce the use of users' eyes. For users who work in front of computers all day, it is another way to receive information in a leisure and relaxing way. Therefore, it is widely welcomed by users. At present, the download volume of some audio novel apps on certain applications exceeds 400 million. With huge user traffic and a very large market, it also shows that audio novel apps have huge commercial value. However, existing users often play audio advertisements once or twice for each episode of audio novels, which seriously affects the user's experience of listening to audio novels. At the same time, the playback of some audio novels only allows the buffering of the currently playing episodes so that the listeners need to listen to the same audio advertisements repeatedly for each episode. This not only wastes user time but also gives a poor user experience.

[0003] However, current Bluetooth headsets directly play all the sounds of mobile terminals. The insertion of audio ads will also affect the listening effect of Bluetooth headsets. It is urgent to switch or remove unnecessary audio ads. The manual inspection method to check whether there are audio ads in the audio played by the application is inefficient, takes a long time to wait, wastes manpower, and has a poor user experience. At the same time, buffering a large amount of audio data through the mobile terminal can easily cause the mobile terminal to freeze. Summary of the invention

[0004] Based on the above situation, the present invention proposes a method for playing audio novels based on cloud space, which plays audio novels through a first application program, obtains the title and chapter information of the audio novel, identifies that the audio novel contains audio advertisements, caches the subsequent chapter audios of the current chapter of the audio novel according to the title and chapter information, and uploads the subsequent chapter audios without audio advertisements to the cloud space, and plays the subsequent chapter audios in the cloud space through a second application program when a characteristic word at the end of a chapter appears in the audio novel. On the one hand, the playback of audio advertisements in subsequent chapters can be reduced, improving the user's listening experience; on the other hand, the amount of data buffered by the user in the mobile terminal can be reduced, avoiding freezes due to excessive buffered data.

[0005] The present invention provides a method for playing audio novels based on cloud space. The method comprises the following steps: playing the audio novel through a first application program, obtaining the title and chapter information of the audio novel; identifying that the audio novel contains audio advertisements; obtaining the subsequent chapter audios of the current chapter of the audio novel according to the title and chapter information, and uploading the subsequent chapter audios excluding the audio advertisements to the cloud space; and playing the subsequent chapter audios in the cloud space through a second application program when a characteristic word at the end of a chapter appears in the audio novel.

[0006] When there is a characteristic word combination in the audio novel, it is determined that the audio novel contains an audio advertisement corresponding to the characteristic word combination. The characteristic word combination includes a first characteristic word, a second characteristic word, and a time difference between the time points when the first characteristic word and the second characteristic word appear.

[0007] The steps for obtaining the feature word combination are: obtaining the first audio and the second audio; comparing the first audio with the second audio; when the comparison result has a comparison audio with the same playback audio content and playback time, identifying the text information in the comparison audio; when there are preset feature words in the comparison audio, determining that the comparison audio is an audio advertisement; and generating a corresponding feature word combination through the audio advertisement.

[0008] When the audio advertisements corresponding to the same characteristic word combination are different, the characteristic word combination of one of the audio advertisements is regenerated.

[0009] At the same time, the present invention also provides an audio novel playing system based on cloud space, comprising: an audio novel information acquisition module, an audio advertisement recognition module, a cloud space cache module and a playing switching module, and each module is signal-connected; The audio novel information acquisition module is used to obtain the title and chapter information of the audio novel when playing the audio novel through the first application; the audio advertisement recognition module is used to identify the audio advertisements in the audio novel; the cloud space cache module is used to obtain the subsequent chapter audio of the current chapter of the audio novel according to the title and chapter information, and upload the subsequent chapter audio excluding the audio advertisement to the cloud space; the playback switching module is used to switch to the second application to play the subsequent chapter audio in the cloud space when the characteristic words at the end of the chapter appear in the audio novel.

[0010] At the same time, the present invention also provides a Bluetooth headset, which is characterized by comprising: a Bluetooth headset body and a charging compartment, and an AI recognition processing module is arranged in the charging compartment to execute to implement the above-mentioned audio novel playing method.

[0011] Some technical effects of the present disclosure are: playing an audio novel through a first program to obtain the title and chapter information of the audio novel; after determining that there are audio advertisements when the audio novel is played, removing the corresponding audio advertisements by manual or system elimination, and then storing the corresponding subsequent chapter audios without audio advertisements in order in the cloud space. By automatically switching to the second program to play the subsequent chapter audios pre-stored in the cloud space, the time users spend listening to audio advertisements is reduced, thereby improving the user experience. At the same time, by continuously acquiring new database content through the system, audio advertisements can be identified more and more efficiently and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] To better understand the technical solution of the present disclosure, you may refer to the following drawings for auxiliary explanation of the prior art or embodiments. These drawings will selectively display the products or methods involved in the prior art or some embodiments of the present disclosure. The basic information of these drawings is as follows: Figure 1 It is a flow chart of an embodiment of a method for playing an audio novel based on cloud space of the present invention; Figure 2 It is a structural schematic diagram of an embodiment of an audio novel playing system based on cloud space of the present invention. DETAILED DESCRIPTION

[0013] The technical means or technical effects involved in the present disclosure will be further described below. Obviously, the embodiments provided are only some of the embodiments of the present disclosure, not all of them. Based on the embodiments in the present disclosure and the explicit or implicit descriptions of the figures and texts, all other embodiments that can be obtained by those skilled in the art without making creative efforts will be within the scope of protection of the present disclosure.

[0014] S1. Play the current audio novel through the first application program, and obtain the title and chapter information of the corresponding audio novel.

[0015] Generally speaking, audio novels have corresponding book titles and chapter information, and several chapters (episodes or rounds) together constitute the complete content of the entire audio novel. For example, "Lao Qian*hu" broadcast by Zi* on the Xima*ya app is divided into 6 chapters, each with more than 300 sections; the specific content corresponding to the audio novel can be found through the chapter information. After the Bluetooth device is connected to the mobile terminal, the book title and chapter information of the corresponding audio will be obtained during playback. For example, when using a car Bluetooth speaker, the corresponding book title and chapter information (including specific chapters and names) can be seen on the car screen. Play the current audio novel through the first application (such as the Xima*ya app) to obtain the corresponding book title and chapter information. Of course, the book title and chapter information corresponding to the audio novel can also be identified through the second application by sound or screen recognition.

[0016] S2. Recognize that the audio novel contains audio advertisements.

[0017] It is recognized that the audio novel currently being played contains audio ads. Audio ads can be identified manually or by the system (AI intelligence). Manual identification mainly involves identifying that the audio novel contains audio ads after a person hears an audio ad; while system (AI intelligence) identification can determine whether an audio ad is contained by comparing empirical data. By playing the current audio novel, the text information in the current audio novel is identified; when there is a characteristic word combination in the current audio novel, it is determined that it contains an audio ad. The specific identification method is as follows: by playing multiple audio novel audios, obtaining multiple comparative audio data, and then identifying whether the audio contains an advertising characteristic word combination by comparing the audio.

[0018] The feature word combination is pre-set in the database, and the database increases the amount of data through repeated AI training. When the corresponding feature word combination appears in the database after the played audio is converted into text, it can be determined that there is an audio advertisement in the audio and the specific audio advertisement and the corresponding playback time are determined according to the database data. After obtaining the audio data, continuous comparison is performed, and the audio advertisement in the audio data is determined by feature word comparison. The advertising feature words and playback time corresponding to each different audio advertisement are continuously obtained, and then the feature word combination corresponding to each audio advertisement is formed through the training model, and the feature word combination is pre-set in the database. The specific method for generating the feature word combination is: obtaining the first audio and the second audio; comparing the first audio with the second audio; when the comparison result has the same playback audio content and playback time in the comparison audio, identifying the text information in the comparison audio; when there is a preset feature word in the comparison audio, the comparison audio is determined to be an audio advertisement. Generate the corresponding feature word combination in the audio advertisement, and pre-set the feature word combination in the database.

[0019] Since different audio novels correspond to different audio advertisements, audio advertisements generally exist at the beginning or end of a round and at the place where the playback is played again after a pause. When the playback time of both audio segments is greater than the playback time of each round, the corresponding audio advertisements can be played for both audio segments. At present, the general time setting for audio advertisements is 10s, 15s, 30s, 60s or even 90s. During the playback process, only by comparing the two audio segments can the system identify whether they are audio advertisements with the same content. Different audio novels and different audio novel apps may produce different audio advertisements. Therefore, there may be multiple results when comparing the first audio and the second audio, which need to be further processed according to the results. Through the system comparison, the corresponding audio data is traversed to find the audio data segments with the same audio content and playback time (the playback time needs to be more than 10s, and if it is too short, it is easy to have multiple high-frequency words that are the same). It is also possible to use AI voice recognition software to convert the voice in the corresponding audio into text, and judge whether the audio content is the same or consistent by the content of the text (because this is related to the function of the voice software, the content and playback time cannot be 100% the same, and it can be judged to be the same if it is higher than a certain percentage). If it is found that there is a comparison audio with the same playback audio content and playback time in the first audio and the second audio, the comparison audio can be converted into the corresponding text through voice recognition software. Currently, audio advertisements involve more related industries such as training, games, dental implants, and financial lending. Their feature words can be associated with their related brand names, functions and related descriptions, and collected into a feature word library. This can improve the efficiency of comparison and reduce the time of text comparison. It is also possible to directly compare the converted text without extracting feature words, but the efficiency is lower.

[0020] Generally speaking, if there is a brand's promotional words or specific promotional terms in a certain audio or video, it can be determined that the audio or video contains a voice advertisement. To determine whether the voice advertisements are the same, it is necessary to determine whether the content and playback time of the voice advertisements are the same. The efficiency of comparing the data of the entire audio segment will be very low. It can be used to determine whether there is a voice advertisement in the audio for the first time, so that enough audio data can be collected and stored in the database. Through the database, the corresponding voice advertisement data is trained on the voice advertisement feature words by AI to generate a feature word combination. The subsequent audio novel can effectively identify the corresponding voice advertisement by identifying the feature word combination. Because just judging that there is a voice advertisement is not enough to efficiently identify which voice advertisement it is, and its playback time cannot be determined, it is impossible to effectively remove or switch to other non-voice advertisement audio. The feature word combination can be a combination of certain words (Qiandu has two feature words Qian and Du to form a combination), or a combination of several phrases; it can also be composed of several words, phrases and time points. If only relying on a single feature word combination to determine whether there is an advertisement, it is easy to mistakenly judge the content of the audio novel as a voice advertisement (for example, a certain brand product is involved in the audio novel). The feature word combination can effectively improve the accuracy of determining the audio advertisement, but the feature word combination is prone to the problem of repeated appearance of the word feature word, which reduces the accuracy of determining the audio advertisement. For example, the first feature word and the second feature word form the first feature word combination, where the second feature word appears twice in the audio, and the determination of the audio advertisement is still determined by the first feature word combination, and different audio advertisements may be determined as the same audio advertisement. Therefore, in order to improve the accuracy of determining the same audio advertisement, in order to obtain an accurate advertisement removal time or switching time. The feature word combination here includes the first feature word appearing at the first time point and the second feature word appearing at the second time point, and the time difference between the first time point and the second time point. When the feature word combination appears in the subsequent audio, it is only necessary to determine whether the first feature word, the second feature word, and whether the same time difference exists to determine that the audio has the audio advertisement, and directly remove the audio advertisement or switch to non-audio advertisement audio according to the corresponding audio advertisement duration. Therefore, the feature word combination includes the first feature word, the second feature word, and the time difference between the time points when the first feature word and the second feature word appear. This embodiment uses the advertising feature word used when determining that the comparison audio is a voice advertisement as the first feature word, and the second feature word can be randomly extracted from the subsequent audio. The time difference between the first feature word and the second feature word is recorded to generate a corresponding feature word combination. Through the comparison of multiple audios and the identification of voice advertisements, the range of feature word combinations can be continuously expanded through AI based on the original feature word library to continuously improve the accuracy of identifying voice advertisements.When the subsequent audio is played, if the same first feature word, second feature word, and time difference between the first feature word and the second feature word appear, it can be determined that there is a specific audio advertisement in the audio, and the system can receive the corresponding audio content according to the time point when the audio advertisement ends. Since the existing audio advertisements are constantly updated, new audio advertisements continue to appear. The feature word combinations generated by two or more audio advertisements may be the same, which will affect the accuracy of the determination of specific audio advertisements. Therefore, the system will compare the new feature word combination with the old feature word combination, and the audio advertisements corresponding to the two are not the same audio advertisements; a new feature word combination will be regenerated. Of course, it can also be an old feature word combination. Here, it is only necessary to make the subsequent feature word combination correspond to the corresponding audio advertisement one by one.

[0021] As a better implementation example, when a voice advertisement is identified, different processing can be performed according to different needs of users. If the user wants to directly remove the advertisement, the system buffer time can be used to cache part of the audio data and re-record it. That is, when it is determined that the third audio has a comparison audio (voice advertisement), the comparison audio is removed from the third audio according to the playback time and the time point when the comparison audio begins to appear in the third audio, and the third audio after the comparison audio is removed is re-recorded to generate the fourth audio. In this way, the fourth audio can remove the voice advertisement and buffer it in the storage area. It is worth mentioning that the removal of voice advertisements is different from the removal of video advertisements. Its data volume will not be particularly large. Basically, the time to regenerate a new audio file after removal can be very short, and the buffering time will not make the user feel uncomfortable. When the user chooses to switch to the audio he likes, the voice advertisement can be replaced with non-advertisement audio when it is determined that the third audio has a comparison audio (voice advertisement).

[0022] When the subsequent audio is played, if the same first feature word, second feature word and the time difference between the appearance time points of the first feature word and the second feature word appear, it can be determined that there is a specific audio advertisement in the audio, and the system can receive the corresponding audio content according to the time point when the audio advertisement ends.

[0023] S3. Obtain the subsequent chapter audio of the current chapter according to the book title and chapter information, generate the subsequent chapter audio without audio advertisements and upload it to the cloud space.

[0024] Through the title and chapter information of the current audio novel, the information of the subsequent chapters can be obtained, and the audio of the subsequent chapters can be downloaded according to the information of the subsequent chapters. Specifically, after obtaining the title and chapter information of the current audio novel, it can be sent to the service center. The service center caches the audio of the subsequent chapters by manually clicking or controlling one or more mobile terminals by a program. For example, 10 mobile terminals can be used to click on the first application and buffer the content of the subsequent 10 sections respectively, and the subsequent chapter audio without audio advertisements in the subsequent 10 sections can be generated by manual elimination or system generation and uploaded to the corresponding cloud space. The method for the system to generate audio without audio advertisements is as follows: when it is determined that there is an audio advertisement in the audio of the subsequent chapter, according to the playback time and the time point when the audio advertisement begins to appear in the third time period, the comparison audio is eliminated from the third audio, and the third audio after eliminating the comparison audio is re-recorded and the fourth audio is generated. In this way, the fourth audio eliminates the audio advertisement and is buffered in the storage area. It is worth mentioning that the elimination of audio ads is different from the elimination of video ads. The amount of data is not particularly large. Basically, the time to regenerate a new audio file after elimination can be very short, and the buffering time will not make the user feel particularly uncomfortable. Collecting the characteristic word combinations found in this embodiment into a database is a reliable method for subsequent efficient identification of various audio ads. At the same time, the cloud space can be replaced by a storable medium. In order not to occupy the hard disk space of the mobile terminal, a storage space can be set in the charging compartment of the Bluetooth headset, and the audio of subsequent chapters can be stored in the storage space of the Bluetooth headset when the headset is used for listening.

[0025] S4. When a chapter ending feature word appears in the audio novel, the subsequent chapter audio in the cloud space is played through the second application.

[0026] When the user reaches the end of each section of the audiobook, the host will have a closing statement to inform the user that the audiobook has ended and the next chapter is needed. By identifying the end-of-chapter feature words said by the host, the second application can be started to play the audio of the subsequent chapters. Generally speaking, the ending feature phrase will appear in the form of "Dear audience, this chapter has ended, welcome to continue listening to the next round". When the system recognizes this keyword phrase, the second application will be started from the background to directly play or obtain the next chapter audio buffered in the cloud space.

[0027] like Figure 2 As shown, this embodiment also includes a cloud-based audio novel playback system, including: an audio novel information acquisition module, an audio advertisement recognition module, a cloud space cache module, and a playback switching module, and each module is signal-connected; The audio novel information acquisition module is used to obtain the title and chapter information of the audio novel when playing the audio novel through the first application; the audio advertisement recognition module is used to identify the audio advertisements in the audio novel; the cloud space cache module is used to cache the audio of subsequent chapters of the current chapter of the audio novel according to the title and chapter information, eliminate the audio advertisements and upload the audio of subsequent chapters to the cloud space; the playback switching module is used to switch to the second application to play the audio of subsequent chapters in the cloud space when the characteristic words at the end of the chapter appear in the audio novel.

[0028] As an application of this embodiment, this embodiment also provides a Bluetooth headset, including: a Bluetooth headset body and a charging compartment, wherein an AI recognition processing module is provided in the charging compartment for implementing the above-mentioned audio novel playing method. At the same time, a storage module is also provided in the charging compartment for storing the audio of subsequent chapters of the audio novel, which is used to replace the storage function of the cloud space (that is, when the characteristic words at the end of a chapter appear in the audio novel, the audio of the subsequent chapters in the storage module is played through the second application). In this way, the audio of the subsequent chapters can be stored in advance when there is a WiFi wireless network, reducing the user's traffic consumption when there is no free network.

[0029] As a simple replacement, this identification processing chip can also be used in other products, such as Bluetooth speakers, etc., which are all simple replacements of this embodiment.

[0030] Those skilled in the art can understand that all or part of the steps in the embodiment can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable medium, which can include a flash drive, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk or an optical disk, and other media that can store program codes. In one embodiment, the present disclosure provides a computer-readable medium, in which a computer program is stored, and the computer program is loaded and executed by a processing module to implement an audio novel playback method.

[0031] Within the scope of the knowledge and ability level of those skilled in the art, the various embodiments or technical features mentioned herein may be combined with each other as other optional embodiments without conflict. These limited number of optional embodiments, which are not listed one by one and are formed by combining a limited number of technical features, still fall within the technical scope disclosed in the present disclosure and can be understood or inferred by those skilled in the art in combination with the drawings and the above text.

[0032] In addition, the description of most embodiments is based on different focuses. For details not described in detail, please refer to the content of the prior art or other relevant descriptions in this article for understanding.

[0033] It is emphasized again that the embodiments listed above are typical and preferred embodiments of the present disclosure, which are only used to explain and interpret the technical solutions of the present disclosure in detail to facilitate the understanding of the readers, and are not intended to limit the protection scope or application of the present disclosure. Any technical solutions obtained by any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for playing audio novels based on cloud space, characterized in that: Playing an audio novel through a first application program to obtain the title and chapter information of the audio novel; identifying that the audio novel contains an audio advertisement; obtaining the audio of subsequent chapters of the current chapter of the audio novel according to the title and chapter information, and uploading the audio of subsequent chapters excluding the audio advertisement to a cloud space; When a chapter ending feature word appears in the audio novel, the subsequent chapter audio in the cloud space is played through the second application.

2. The audio novel playing method according to claim 1, characterized in that: When there is a characteristic word combination in the audio novel, it is determined that the audio novel contains an audio advertisement corresponding to the characteristic word combination.

3. The audio novel playing method according to claim 2, characterized in that: The feature word combination includes a first feature word, a second feature word, and a time difference between the time points at which the first feature word and the second feature word appear.

4. The audio novel playing method according to claim 2 or 3, characterized in that: The steps for obtaining the feature word combination are: obtaining the first audio and the second audio; comparing the first audio with the second audio; when the comparison result has a comparison audio with the same playback audio content and playback duration, identifying the text information in the comparison audio; when there is a preset feature word in the comparison audio, determining that the comparison audio is an audio advertisement; and generating a corresponding feature word combination through the audio advertisement.

5. The audio novel playing method according to claim 4, characterized in that: When the audio advertisements corresponding to the same characteristic word combination are different, the characteristic word combination of one of the audio advertisements is regenerated.

6. A cloud-based audio novel playback system, characterized in that: include: Audio novel information acquisition module, audio advertisement recognition module, cloud space cache module and playback switching module, and signal connection of each module; The audio novel information acquisition module is used to acquire the title and chapter information of the audio novel when the audio novel is played through the first application program; the audio advertisement recognition module is used to recognize the audio advertisement in the audio novel; The cloud space cache module is used to obtain the audio of subsequent chapters of the current chapter of the audio novel according to the book title and chapter information, and upload the audio of the subsequent chapters excluding audio advertisements to the cloud space; the playback switching module is used to switch to the second application to play the audio of the subsequent chapters in the cloud space when the characteristic words at the end of the chapter appear in the audio novel.

7. The audio novel playback system according to claim 6, characterized in that: When there is a characteristic word combination in the audio novel, it is determined that the audio novel contains an audio advertisement corresponding to the characteristic word combination; the characteristic word combination includes a first characteristic word, a second characteristic word and a time difference between the time points when the first characteristic word and the second characteristic word appear.

8. The audio novel playback system according to claim 7, characterized in that: The steps for obtaining the feature word combination are: obtaining the first audio and the second audio; comparing the first audio with the second audio; when the comparison result has a comparison audio with the same playback audio content and playback duration, identifying the text information in the comparison audio; when there is a preset feature word in the comparison audio, determining that the comparison audio is an audio advertisement; and generating a corresponding feature word combination through the audio advertisement.

9. A Bluetooth headset, characterized in that Comprising: comprising: a Bluetooth headset body and a charging compartment, wherein an AI recognition processing module is arranged in the charging compartment for implementing the audio novel playing method described in any one of claims 1-5.

10. The Bluetooth headset according to claim 9, characterized in that: A storage module is also provided in the charging compartment to replace the cloud space to store the subsequent chapter audio of the audio novel. When the characteristic words at the end of a chapter appear in the audio novel, the subsequent chapter audio in the storage module is played through the second application.

Images