Metadata generation methods, identification methods, and electronic devices

Abstract

This application provides a method for generating and identifying metadata, as well as an electronic device. The method includes: first, acquiring media content; then, generating metadata for the media content, the metadata including an AIGC (Artificial Intelligence Generated Content) identifier, which indicates whether the media content is AIGC. Subsequently, the AIGC identifier in the media content's metadata can be used to identify whether the media content is AIGC; this can help users determine the authenticity of media content to a certain extent, and increase users' vigilance when using (e.g., forwarding) media content.

Description

Technical Field

[0001] This application relates to the field of data processing, and more particularly to a method for generating and identifying metadata, and an electronic device. Background Technology

[0002] With the development of artificial intelligence (AI) technology, AI-generated content (AIGC) has experienced explosive growth. AIGC can create novel, interesting, and diverse digital content (such as images, music, and natural language) to meet the preferences and scenarios of different users.

[0003] However, AI-generated content may affect the authenticity and credibility of information; for example, AI-generated content could be used to spread misinformation. Therefore, identifying AI-generated content is crucial. Summary of the Invention

[0004] To help users identify whether media content is AIGC, this application provides a metadata generation method, an identification method, and an electronic device. The metadata generated by this method can include an AIGC identifier indicating whether the media content is AIGC. Subsequently, the AIGC identifier in the media content's metadata can be used to identify whether the media content is AIGC. This can help users determine the authenticity of media content to a certain extent and increase their vigilance when using (e.g., forwarding) media content.

[0005] In a first aspect, embodiments of this application provide a method for generating metadata. The method includes: first, acquiring media content; then, generating metadata of the media content, wherein the metadata includes an AIGC (Artificial Intelligence Generated Content) identifier, the AIGC identifier indicating whether the media content is AIGC. Subsequently, the AIGC identifier can be obtained from the metadata of the media content, and the AIGC identifier can be used to determine whether the media content is AIGC. This approach can help users determine the authenticity of media content to some extent (for example, when media content is determined to be AIGC, it can be assumed that the information contained in the media content is highly likely to be false, and further verification can be conducted using other methods). It also increases users' vigilance when using (e.g., forwarding) media content (for example, when media content is determined to be AIGC, users can further verify whether the media content is copyrighted, whether it contains personal information, and whether it contains false information, etc. If the media content is determined to be copyrighted, contains personal information, or contains false information, users can choose not to use the media content. If the media content is verified to be uncopyrighted, does not contain personal information, and does not contain false information, users can use the media content).

[0006] For example, media content includes, but is not limited to, images, graphics, videos, and audio, and this application does not limit this.

[0007] For example, metadata for media content may include data used to describe the media content; for instance, metadata may include information describing media content properties.

[0008] It should be noted that the metadata generation method of this application is applied to the creation stage of media assets, corresponding to the metadata generation process in the media asset creation stage.

[0009] For example, when the media content is AIGC, the AIGC identifier in the media content's metadata is a preset identifier value, indicating that the media content is AIGC. When the media content is not AIGC, the AIGC identifier in the media content's metadata is another identifier value, indicating that the media content is not AIGC. It should be understood that the preset identifier value is different from the other identifier values; for the AIGC identifier, one preset identifier value and one or more other identifier values ​​can be predefined.

[0010] According to the first aspect, the metadata of the media content also includes a first hash value, which is the hash value of the data including the AIGC identifier. This prevents the AIGC identifier from being tampered with.

[0011] For example, the first hash value is the hash value of data including the AIGC identifier, which can be understood as the first hash value being the hash value of some data; wherein, "some data" may include the AIGC identifier.

[0012] For example, when "some data" only includes the AIGC identifier, the first hash value can refer to the hash value of the AIGC identifier, or in other words, the first hash value is obtained by hashing the AIGC identifier. Specifically, a hash algorithm can be used to hash the value of the AIGC identifier to obtain the first hash value. It should be understood that this application does not limit the hash algorithm used to calculate the first hash value.

[0013] It should be understood that the first hash value is optional information in the metadata of the media content.

[0014] According to the first aspect, or any implementation of the first aspect above, the metadata of the media content also includes a digital signature of the first hash value. This prevents the AIGC identifier from being forged.

[0015] For example, a digital signature of the first hash value can be obtained by digitally signing the first hash value using a private key based on a digital signature algorithm. It should be understood that this application does not limit the digital signature algorithm used to calculate the first hash value.

[0016] It should be understood that the digital signature of the first hash value is optional information in the metadata of the media content.

[0017] According to the first aspect, or any implementation of the first aspect above, the metadata of the media content also includes a training enable flag, which indicates whether the media content is allowed to be used for model training. This facilitates subsequent determination of whether the media content can be used for AI model training.

[0018] For example, the training enable flag can be the DNT (DonotTrain) field.

[0019] It should be understood that training enablement is identified as optional information in the metadata of the media content.

[0020] According to the first aspect, or any of the implementations of the first aspect above, the metadata of media content also includes relevant information about the media content.

[0021] For example, information related to media content can refer to information related to the media content. This information may include data used to describe the media content, such as information used to describe the attributes of the media content.

[0022] According to the first aspect, or any implementation of the first aspect above, the metadata of the media content also includes at least one of the following: model identifier, model version identifier, model parameter data volume identifier, or model training set data volume identifier.

[0023] For example, model identifiers can indicate the type of AI model that generates media content, such as ChatGPT (an AI-powered natural language processing tool), Midjourney (an AI drawing tool), etc.

[0024] For example, a model version identifier can indicate the version of the AI ​​model that generates the media content; such as MidjourneyV4, MidjourneyV5, ChatGPTv3.5, ChatGPTv4, etc.

[0025] For example, the model parameter data volume identifier can indicate the size of the AI ​​model parameters that generate media content, such as 1 million parameters or 10 million parameters.

[0026] For example, a model training set data size identifier can indicate the size of the training set data for the AI ​​model used to generate media content, such as 1G or 47T.

[0027] This makes it easier to obtain relevant information about the AI ​​model that generated the media content based on the media content's metadata.

[0028] It should be understood that model identifier, model version identifier, model parameter data volume identifier, and model training set data volume identifier are all optional information in the metadata of media content.

[0029] According to the first aspect, or any implementation of the first aspect above, the metadata of the media content also includes: a second hash value, a third hash value, and a digital signature of the third hash value. The second hash value is the hash value of the media content, and the third hash value is the hash value of data including information related to the media content and the second hash value. This prevents the media content and its related information from being tampered with, and also prevents the related information from being forged.

[0030] For example, the third hash value is the hash value of data that includes information related to the media content and the second hash value. It can be understood that the third hash value is the hash value of certain data, where "certain data" may include information related to the media content and the second hash value.

[0031] For example, when "some data" only includes information related to the media content and a second hash value, the third hash value can be obtained by hashing the information related to the media content and the second hash value. Specifically, the third hash value can be obtained by hashing the information related to the concatenated media content and the second hash value.

[0032] It should be understood that the digital signatures of the second, third, and fourth hash values ​​are all optional information in the metadata of the media content.

[0033] According to the first aspect, or any implementation of the first aspect above, the metadata of the media content also includes one or more media claims and one or more fourth hash values; the media claim is a claim on the media content, and the fourth hash value is the hash value of the media claim.

[0034] For example, media statements may include, but are not limited to: copyright information, thumbnails, whether they can be modified, operating permissions, usage permissions, the source of the media content, editorial information, etc.

[0035] It should be understood that media claims and fourth hash values ​​are optional information in metadata. In other words, media content can be claimed during the metadata generation process, or it can be left unclaimed. This improves the flexibility and backward compatibility of metadata.

[0036] According to the first aspect, or any implementation of the first aspect above, the metadata of the media content also includes a fifth hash value, which is a hash value of data including the second hash value, the third hash value, the digital signature of the third hash value, and one or more fourth hash values. This allows multiple hash values ​​to be associated, further preventing the media content and related information from being tampered with.

[0037] For example, the fifth hash value is the hash value of data that includes the second hash value, the third hash value, the digital signature of the third hash value, and one or more fourth hash values. It can be understood that the fifth hash value is the hash value of certain data, wherein "certain data" may include: the second hash value, the third hash value, the digital signature of the third hash value, and one or more fourth hash values.

[0038] For example, when "some data" only includes a second hash value, a third hash value, a digital signature of the third hash value, and one or more fourth hash values, the fifth hash value may be obtained by hashing the second hash value, the third hash value, the digital signature of the third hash value, and one or more fourth hash values.

[0039] For example, an intermediate hash value can be obtained by concatenating the second hash value, the third hash value, the digital signature of the third hash value, and one or more of the fourth hash values. If the intermediate hash value is obtained by concatenating all items from the second hash value, the third hash value, the digital signature of the third hash value, and the fourth hash value, then the intermediate hash value can be used as the fifth hash value. If the intermediate hash value is obtained by concatenating some items from the second hash value, the third hash value, the digital signature of the third hash value, and some of the fourth hash values, then the fifth hash value can be obtained by hashing the intermediate hash value with another part of the second hash value, the third hash value, the digital signature of the third hash value, and one or more of the fourth hash values.

[0040] For example, the metadata of media content may also include a digital signature of a fifth hash value; it should be understood that the metadata of media content may also include other information, which is not limited in this application.

[0041] Based on the first aspect, or any of the above implementations of the first aspect, the media assets corresponding to the media content are obtained based on the media content and the media content's metadata.

[0042] For example, media content and its metadata can be encapsulated to obtain media assets corresponding to the media content. For instance, if the media content is a JPEG (Joint Picture Encoding Group) image, encapsulating the image and its metadata can yield a JPEG file (i.e., a media asset).

[0043] According to the first aspect, or any implementation of the first aspect above, the metadata of the media content is a JUMBF (JPEG Universal Metadata Box Format) box. The JUMBF box includes a JUMBF description box and a JUMBF content box, and the JUMBF description box and / or the JUMBF content box includes an AIGC identifier. Thus, whether the media content is AIGC can be identified based on the AIGC identifier read from the JUMBF description box and / or the JUMBF content box.

[0044] For example, the JUMBF description box is called the JUMBF Description Box, and the JUMBF content box is called the JUMBF Content Box.

[0045] For example, a new content type (ContentType) can be defined in JUMBF, which can be called AIGCContentType, and the corresponding JUMBF Content Box can be called AICG Content Box. Furthermore, a new identifier value (which can be called a preset identifier value) can be defined for the TYPE field in the JUMBFDescriptionBox. The preset identifier value can indicate that the JUMBF Content Box is an AICG Content Box (or indicate that the type of the JUMBF Content Box is AIGC Content Type, which can also be understood as indicating that the media content is AIGC). In this case, the AIGC identifier can refer to the TYPE field in the JUMBF Description Box.

[0046] For example, this application can add a new field AIGCI (AIGCIidentify) to the JUMBFDescriptionBox. The AIGCI field can indicate whether the media content is AIGC; the AIGCI field in the JUMBFDescription Box can be used as an AIGC identifier. For example, multiple identifier values ​​can be defined for the AIGCI field (including a preset identifier value and at least one other identifier value); for example, two identifier values ​​"1" and "0" can be defined. "1" can be the preset identifier value, indicating that the media content is AIGC, and "0" can be another identifier value, indicating that the media content is not AIGC.

[0047] For example, a new field AIGCI can be added to JUMBFContentBox. The AIGCI field can indicate whether the media content is AIGC; the AIGCI field in JUMBFContentBox can be used as an AIGC identifier.

[0048] For example, JUMBFDescriptionBox may also include a training enable flag.

[0049] According to the first aspect, or any implementation of the first aspect above, the JUMBF content box also includes information related to the media content.

[0050] According to the first aspect, or any implementation of the first aspect above, the JUMBF content box also includes at least one of the following: model identifier, model version identifier, model parameter data volume identifier, or model training set data volume identifier.

[0051] When the media content is AIGC, JUMBFContentBox can be called AICGContentBox.

[0052] For example, AICGContentBox may include, but is not limited to: AIGCI field, AIModel (AI model) field, AIModelVer (AI Model Version) field, ParaSize (ParameterSize, AI model parameter data size) field, DataSize (AI model training set data size) field, and Data (load data) field.

[0053] For example, the AIGCI field (also known as the AIGC identifier, where the AIGC identifier represented by the TYPE field in JUMBFDescriptionBox and the AIGC identifier represented by the AIGCI field in AICGContentBox are two different identifiers, but they serve the same purpose) can also indicate whether JUMBFContentBox is an AICGContent Box (or, in other words, whether the media content is AIGC). For instance, two identifier values ​​are defined for the AIGCI field: "1" and "0"; "1" indicates that JUMBFContentBox is an AICGContentBox, and "0" indicates that JUMBFContentBox is not an AICGContentBox.

[0054] For example, the AIModel field can also be called the model identifier.

[0055] For example, the AIModelVer field can also be called the model version identifier.

[0056] For example, the ParaSize field can also be called the model parameter data volume identifier.

[0057] For example, the DataSize field can also be referred to as the data volume identifier for the model training set.

[0058] For example, the Data field can be used to store information related to media content.

[0059] For example, the JUMBF content box also includes a second hash value, a third hash value, and a digital signature of the third hash value.

[0060] For example, the JUMBF content box also includes one or more media claims and one or more fourth hash values.

[0061] For example, the JUMBF content box also includes a fifth hash value.

[0062] For example, the JUMBF content box also includes a training enable flag.

[0063] It should be understood that the training enable identifier is optional information in the metadata of the media content, and this application does not restrict whether the training enable identifier is located in JUMBFContentBox or JUMBFDescriptionBox.

[0064] According to the first aspect, or any implementation of the first aspect above, the relevant information of the media content includes at least one of the following: the generation time of the media content, the author's name of the media content, the digital content identifier of the media content, the generation location of the media content, the information of the media content generation device, the resolution of the media content, or the size of the media content.

[0065] It should be understood that information related to media content may also include other information such as the media type or the method of media content generation, and this application does not limit this.

[0066] For example, media types may include, but are not limited to: image types, video types, audio types, graphic types, and text types.

[0067] For example, media content can be generated in the following ways: AI generation and non-AI generation. For instance, if the media content is an image, the corresponding generation methods may include, but are not limited to: image generation, screenshot generation, and AI generation; among these, image generation and screenshot generation are non-AI generation methods.

[0068] In this way, in addition to identifying whether the media content is AIGC based on the AIGC identifier read from JUMBFDescriptionBox or JUMBFContentBox, it is also possible to identify whether the media content is AIGC based on the generation method of the media content read from the Data field of JUMBFContentBox.

[0069] According to the first aspect, or any implementation of the first aspect above, the JUMBF descriptor also includes a first hash value, an enable status identifier, a content request identifier, an identity identifier, and a sixth hash value. The sixth hash value is determined based on the JUMBF content frame. The first hash value is the hash value of data including the AIGC identifier, the enable status identifier, the content request identifier, the identity identifier, and the sixth hash value. In other words, the field values ​​of each field in the JUMBF descriptor can be hashed to obtain the first hash value; this prevents the field value of any field in the JUMBF descriptor from being tampered with.

[0070] For example, the first hash value is the hash value of data including AIGC identifier, enable status identifier, content request identifier, identity identifier and sixth hash value. It can be understood that the first hash refers to the hash value of certain data, where "certain data" may include AIGC identifier, enable status identifier, content request identifier, identity identifier and sixth hash value.

[0071] For example, when "some data" only includes the hash values ​​of the AIGC identifier, the enable status identifier, the content request identifier, the identity identifier, and the sixth hash value, the first hash value can be obtained by hashing the AIGC identifier, the enable status identifier, the content request identifier, the identity identifier, and the sixth hash value.

[0072] For example, an intermediate hash value can be obtained by concatenating one or more of the following: the AIGC identifier, the enable status identifier, the content request identifier, the identity identifier, and the sixth hash value. If the intermediate hash value is obtained by concatenating all items from the AIGC identifier, enable status identifier, content request identifier, identity identifier, and the sixth hash value, then the intermediate hash value can be used as the first hash value. If the intermediate hash value is obtained by concatenating some items from the AIGC identifier, enable status identifier, content request identifier, identity identifier, and the sixth hash value, then the first hash value can be obtained by hashing the intermediate hash value with another part of these three items.

[0073] For example, JUMBFDescriptionBox may include multiple fields: TYPE field (type field), TOGGLES(T) field (enabled status field), LABEL field (content request field), ID field (identity field), and SHA256HASH field (field used to store the sixth hash value).

[0074] For example, the TYPE field (which can be called the JUMBF type field or type identifier) ​​can be used to describe the JUMBF content type (that is, the type of the JUMBF content box).

[0075] For example, the TYPE field can contain a 16-byte UUID whose value specifies the type of the JUMBFContent Box.

[0076] For example, the types of JUMBF Content Boxes can include, but are not limited to: Codestream ContentType, UUID Content Type, XML Content Type, JSON Content Type, CBOR Content Type, Embedded File Content Type, and other content types (which can be customized).

[0077] For example, the TYPE field is a required field.

[0078] For example, the TOGGLES(T) field (also known as the enable status flag) can contain a 1-byte (i.e., 8-bit) binary number that can be used to represent the value of an option associated with the JUMBF box. Each bit represents an option; a value of 1 indicates that the option is enabled, and a value of 0 indicates that the option is disabled.

[0079] For example, the LABEL field (also known as a content request identifier) ​​can contain a variable-length text label that can be used to reference or request the content of the JUMBF box.

[0080] For example, the ID field (also known as the identity identifier) ​​can contain a unique 4-byte ID assigned by the user, which can be used to reference the JUMBF box in binary.

[0081] For example, the SHA256HASH field can be used to store a sixth hash value, which is determined based on n JUMBF Content Boxes. For example, the sixth hash value can be obtained by concatenating the n JUMBF Content Boxes in their order within the JUMBFSuper Box and then performing a hash calculation.

[0082] For example, the TOGGLES(T) field, LABEL field, ID field, and SHA256HASH field can be optional fields.

[0083] For example, a hash algorithm can be used to calculate the value of at least one of the following fields: the identifier value of the concatenated AIGC identifier, the TOGGLES(T) field, the LABEL field, the ID field, the SHA256HASH field, and the DNT field, to obtain the first hash value.

[0084] For example, the JUMBF descriptor box may also include a digital signature of the first hash value.

[0085] Secondly, embodiments of this application provide an identification method, which includes: first, obtaining metadata of media content; then, reading an AIGC identifier from the metadata of the media content; and finally, determining whether the media content is AIGC based on the AIGC identifier.

[0086] According to the second aspect, when the value of the AIGC identifier is a preset identifier value, the media content is determined to be AIGC.

[0087] According to the second aspect, or any implementation of the second aspect above, the metadata of the media content is a JPEG Universal Metadata Frame (JUMBF) frame, which includes a JUMBF description frame and a JUMBF content frame, and the JUMBF description frame and / or the JUMBF content frame includes an AIGC identifier.

[0088] Thirdly, embodiments of this application propose a metadata generation apparatus, which includes: a content acquisition module for acquiring media content; and a metadata generation module for generating metadata of the media content, wherein the metadata of the media content includes an AIGC identifier, and the AIGC identifier indicates whether the media content is AIGC.

[0089] It should be understood that the metadata generation device of the third aspect can perform the steps in the first aspect and any implementation thereof, which will not be elaborated here.

[0090] The third aspect and any implementation thereof correspond to the first aspect and any implementation thereof, respectively. The technical effects of the third aspect and any implementation thereof are similar to those of the first aspect and any implementation thereof, and will not be repeated here.

[0091] Fourthly, embodiments of this application propose an identification device, which includes: a metadata acquisition module for acquiring metadata of media content; a reading module for reading an AIGC identifier from the metadata of the media content; and a judgment module for judging whether the media content is AIGC based on the AIGC identifier.

[0092] It should be understood that the identification device of the fourth aspect can perform the steps in the second aspect and any implementation thereof, which will not be elaborated here.

[0093] The fourth aspect and any implementation thereof correspond to the second aspect and any implementation thereof, respectively. The technical effects of the fourth aspect and any implementation thereof can be found in the technical effects of the second aspect and any implementation thereof, as described above, and will not be repeated here.

[0094] Fifthly, embodiments of this application provide an electronic device, including: a memory and a processor, the memory being coupled to the processor; the memory storing program instructions, which, when executed by the processor, cause the electronic device to perform the metadata generation method in the first aspect or any possible implementation of the first aspect.

[0095] The fifth aspect and any implementation thereof correspond to the first aspect and any implementation thereof, respectively. The technical effects of the fifth aspect and any implementation thereof are similar to those of the first aspect and any implementation thereof, and will not be repeated here.

[0096] In a sixth aspect, embodiments of this application provide an electronic device, including: a memory and a processor, the memory being coupled to the processor; the memory storing program instructions, when executed by the processor, causing the electronic device to perform the identification method in the second aspect or any possible implementation of the second aspect.

[0097] The sixth aspect and any implementation thereof correspond to the second aspect and any implementation thereof, respectively. The technical effects of the sixth aspect and any implementation thereof are similar to those of the second aspect and any implementation thereof, and will not be repeated here.

[0098] In a seventh aspect, embodiments of this application provide a chip including one or more interface circuits and one or more processors; the interface circuits are configured to receive signals from the memory of an electronic device and send signals to the processors, the signals including computer instructions stored in the memory; when the processor executes the computer instructions, the steps in the first aspect or any possible implementation of the first aspect are executed.

[0099] The seventh aspect and any implementation thereof correspond to the first aspect and any implementation thereof, respectively. The technical effects of the seventh aspect and any implementation thereof are similar to those of the first aspect and any implementation thereof, and will not be repeated here.

[0100] Eighthly, embodiments of this application provide a chip including one or more interface circuits and one or more processors; the one or more processors receive or send data through the one or more interface circuits, and when the one or more processors execute computer instructions, the steps in the second aspect or any possible implementation of the second aspect are executed.

[0101] The eighth aspect and any implementation thereof correspond to the second aspect and any implementation thereof, respectively. The technical effects corresponding to the eighth aspect and any implementation thereof are similar to those corresponding to the second aspect and any implementation thereof, and will not be repeated here.

[0102] Ninthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when run on a computer or processor, causes the computer or processor to perform the metadata generation method in the first aspect or any possible implementation thereof.

[0103] The ninth aspect and any implementation thereof correspond to the first aspect and any implementation thereof, respectively. The technical effects corresponding to the ninth aspect and any implementation thereof are similar to those corresponding to the first aspect and any implementation thereof, and will not be repeated here.

[0104] In a tenth aspect, embodiments of this application provide a computer-readable storage medium storing a computer program that, when run on a computer or processor, causes the computer or processor to perform the identification method in the second aspect or any possible implementation thereof.

[0105] The tenth aspect and any implementation thereof correspond to the second aspect and any implementation thereof, respectively. The technical effects corresponding to the tenth aspect and any implementation thereof are similar to those corresponding to the second aspect and any implementation thereof, and will not be repeated here.

[0106] Eleventhly, embodiments of this application provide a computer program product, which includes computer instructions that, when executed by a computer or processor, cause the computer or processor to perform a metadata generation method in the first aspect or any possible implementation thereof.

[0107] The eleventh aspect and any implementation thereof correspond to the first aspect and any implementation thereof, respectively. The technical effects corresponding to the eleventh aspect and any implementation thereof can be found in the technical effects corresponding to the first aspect and any implementation thereof, as described above, and will not be repeated here.

[0108] In a twelfth aspect, embodiments of this application provide a computer program product, the computer program product including computer instructions, which, when executed by a computer or processor, cause the computer or processor to perform the identification method in the second aspect or any possible implementation of the second aspect.

[0109] The twelfth aspect and any implementation thereof correspond to the second aspect and any implementation thereof, respectively. The technical effects corresponding to the twelfth aspect and any implementation thereof are similar to those corresponding to the second aspect and any implementation thereof, and will not be repeated here. Attached Figure Description

[0110] Figure 1a This is a schematic diagram illustrating an application scenario;

[0111] Figure 1b This is a schematic diagram illustrating an application scenario;

[0112] Figure 1c A schematic diagram illustrating an exemplary identification method;

[0113] Figure 2 This is a schematic diagram illustrating the metadata generation process as an example.

[0114] Figure 3a This is a schematic diagram of the structure of a JUMBF frame as an example.

[0115] Figure 3b This is a schematic diagram of the structure of the JUMBFDescriptionBox as an example.

[0116] Figure 3c This is a schematic diagram of the structure of the AICGContentBox as an example.

[0117] Figure 4a This is a schematic diagram illustrating the metadata generation process as an example.

[0118] Figure 4b This is a schematic diagram of the structure of a JUMBF frame as an example.

[0119] Figure 5a This is a schematic diagram of the structure of the JUMBFDescriptionBox as an example.

[0120] Figure 5b This is a schematic diagram illustrating the metadata generation process as an example.

[0121] Figure 5c This is a schematic diagram of the structure of the JUMBFContentBox as an example.

[0122] Figure 5d This is a schematic diagram of the structure of a JUMBF frame as an example.

[0123] Figure 6a This is a schematic diagram of the structure of the JUMBFContentBox as an example.

[0124] Figure 6b This is a schematic diagram illustrating the metadata generation process as an example.

[0125] Figure 6c This is a schematic diagram of the structure of a JUMBF frame as an example.

[0126] Figure 7a This is a schematic diagram of the structure of the JUMBFDescriptionBox as an example.

[0127] Figure 7b This is a schematic diagram of the structure of AIGCContentBox as an example;

[0128] Figure 7c This is a schematic diagram of the structure of the JUMBFDescriptionBox as an example.

[0129] Figure 7d This is a schematic diagram of the structure of the JUMBFDescriptionBox as an example.

[0130] Figure 7e This is a schematic diagram illustrating the structure of the Data field in a JUMBFContentBox;

[0131] Figure 7f This is a schematic diagram illustrating the fifth hash value generation process as an example.

[0132] Figure 7g This is a schematic diagram illustrating the fifth hash value generation process as an example.

[0133] Figure 8 This is a schematic diagram illustrating the identification process as an example.

[0134] Figure 9 This is a schematic diagram illustrating the identification process as an example.

[0135] Figure 10 This is a schematic diagram illustrating the identification process as an example.

[0136] Figure 11 This is a schematic diagram illustrating the identification process as an example.

[0137] Figure 12 A schematic diagram of a metadata generation apparatus as an example;

[0138] Figure 13 This is a schematic diagram of an identification device as an example.

[0139] Figure 14 This is a schematic diagram of the structure of an exemplary device. Detailed Implementation

[0140] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0141] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0142] The terms "first" and "second," etc., used in the specification and claims of this application are used to distinguish different objects, not to describe a specific order of objects. For example, "first target object" and "second target object," etc., are used to distinguish different target objects, not to describe a specific order of target objects.

[0143] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0144] In the description of the embodiments in this application, unless otherwise stated, "multiple" means two or more. For example, multiple processing units means two or more processing units; multiple systems means two or more systems.

[0145] To facilitate understanding, some technical terms mentioned in the embodiments of this application will be introduced first.

[0146] 1. Media Assets

[0147] Media assets can include media content and its metadata. The metadata is bound to the media content, or in other words, it is associated with it.

[0148] It should be noted that in some scenarios, media assets are also referred to as media files; for example, if the media content is a JPEG (Joint Picture Encoding Group) image, the corresponding media asset can be called a JPEG file.

[0149] 2. Media Content

[0150] Media content, also known as digital content, refers to various types of content such as text, images, and sound that exist in digital form. It can be stored on digital media such as CDs and hard drives and disseminated through networks and other means. Digital content encompasses products or services that integrate and apply images, text, audio, and video content using digital technology; it is a product of the combination of digital media technology and cultural creativity.

[0151] Digital technology is a science and technology that has developed alongside electronic computers. It refers to the technology of converting various types of information, including images, text, sound, and video, into binary digits "0" and "1" that electronic computers can recognize, and then performing calculations, processing, storage, transmission, dissemination, and reconstruction. Because computers are needed to encode, compress, and decode information during the calculation and storage processes, it is also called digital technology, computer digital technology, etc. Digital technology is also known as digital control technology.

[0152] 3. Metadata

[0153] Metadata can include data used to describe media content; for example, metadata can include information describing media content properties.

[0154] Figure 1a This is a schematic diagram illustrating an exemplary application scenario. Figure 1a In this context, the media content consists of images.

[0155] Reference Figure 1a For example, after user A's phone A receives an image sent by user B's phone B, if user A needs to use the image (e.g., forward it), phone A can first determine whether the image is AIGC (Artificial Intelligence of Things). After determining that the image is AIGC, it can further verify whether the media content has copyright, determine whether the media content contains personal information, and verify whether the media content contains false information, etc.; and then determine whether to use the image.

[0156] It should be understood that, Figure 1a The mobile phone A mentioned can also be other electronic devices with strong computing power, such as personal computers, computer workstations, tablet computers, etc. This application does not limit this.

[0157] Figure 1b This is a schematic diagram illustrating an exemplary application scenario. Figure 1b In this context, the media content consists of images.

[0158] Reference Figure 1b For example, when a user needs to determine whether an image in a smartwatch is AIGC, the image in the smartwatch can be sent to a mobile phone, which will then determine whether the image is AIGC.

[0159] It should be understood that, Figure 1b The smartwatch mentioned can also be other electronic devices with weaker computing power, such as media consumption devices, wearable devices, set-top boxes, game consoles, etc., and this application does not impose any restrictions on this. Figure 1b The mobile phone in this application can also be other electronic devices with strong computing power, such as personal computers, computer workstations, tablet computers, etc. This application does not impose any restrictions on this.

[0160] Figure 1c This is a schematic diagram illustrating an exemplary identification method. Figure 1c The image shows the process of using a mobile phone to determine whether an image is AIGC.

[0161] For example, users can open the AIGC recognition platform client (such as an application, mini-program, webpage, etc.) on their mobile phones to perform AIGC recognition.

[0162] Reference Figure 1c (1) The main interface 101 of the AIGC recognition platform client may include one or more controls, including but not limited to: input box, recognition button, etc. This application does not limit this.

[0163] For example, a user can click the input box, and the AIGC recognition platform's client can respond to the user's action by displaying a file selection interface. Then, after selecting the image to be recognized in the file selection interface, the user can click the recognition button on the main interface 101. Correspondingly, the AIGC recognition platform's client can respond to the user's action by obtaining the image's metadata (it should be understood that when phone A receives an image sent by phone B (or when a phone receives an image sent by a smartwatch), it also receives the image's metadata). Then, based on the image's metadata, it identifies whether the image is AIGC.

[0164] In one possible approach, the AIGC recognition platform's client can perform the recognition of whether an image is AIGC. For example, when there is only one image to be recognized, the AIGC recognition platform's client can perform the recognition locally.

[0165] In one possible approach, the AIGC recognition platform's server can perform the identification of whether an image is AIGC. For example, when there are multiple images to be identified, the AIGC recognition platform's client can send these images to the AIGC recognition platform's server, which will then identify the images and return the results to the client.

[0166] It should be noted that in this application, whether the recognition of whether an image is AIGC is performed by the client of the AIGC recognition platform or by the server of the AIGC recognition platform, is not limited by the number of images to be recognized.

[0167] For example, the client of the AIGC recognition platform can display the recognition results. These results can include whether the image is AIGC or not.

[0168] Reference Figure 1c (2), Figure 1c (2) shows the recognition result as AIGC image.

[0169] Reference Figure 1c (3), Figure 1c The recognition result shown in (3) is that the image is not AIGC.

[0170] It should be understood that users can also manually view the image's metadata and then identify whether the image is AIGC based on the image's metadata; this application does not impose any restrictions on this.

[0171] The following explains the process of generating metadata and identifying media content.

[0172] First, the process of generating metadata for media content will be explained.

[0173] Figure 2 This is a schematic diagram illustrating the metadata generation process as an example.

[0174] S201, Obtain Media Content.

[0175] For example, media content may include, but is not limited to, images, audio, video, text, and graphics, etc., and this application does not limit it.

[0176] For example, while acquiring media content, related information about the media content can also be acquired. This related information can refer to information associated with the media content and may include data describing the media content, such as information describing the attributes of the media content. For example, the related information may include, but is not limited to: the generation time of the media content, the author's name, the digital content identifier, the generation location, the generation device information, the resolution, the size, the media type, or the generation method of the media content.

[0177] For example, media types may include, but are not limited to: image types, video types, audio types, graphic types, and text types.

[0178] For example, media content can be generated in the following ways: AI generation and non-AI generation. For instance, if the media content is an image, the corresponding generation methods may include, but are not limited to: image generation, screenshot generation, and AI generation; among these, image generation and screenshot generation are non-AI generation methods.

[0179] For example, the media content in S201 may refer to media content that does not have metadata. Metadata of the media content can be generated based on the relevant information of the media content, i.e., S202 is executed.

[0180] It should be noted that the metadata generation method of this application is applied to the creation stage of media assets, corresponding to the metadata generation process in the media asset creation stage.

[0181] S202, Generate metadata for media content. The metadata for media content includes an AIGC identifier, which indicates whether the media content is AIGC.

[0182] For example, during the generation of media content metadata, when the media content is determined to be AIGC, the value of the AIGC identifier can be set to a preset identifier value; where the preset identifier value indicates that the media content is AIGC. When the media content is determined not to be AIGC, the value of the AIGC identifier can be set to another identifier value; where the other identifier value indicates that the media content is not AIGC. Next, the media content metadata is generated using the relevant information of the media content and the AIGC identifier; that is, the media content metadata can include the relevant information of the media content and the AIGC identifier.

[0183] For example, one can determine whether media content is AIGC based on the generation method of the media content in its related information. When the media content is generated using AI, it can be determined that the media content is AIGC; when the media content is generated using a non-AI generation method, it can be determined that the media content is not AIGC.

[0184] In this way, the AIGC identifier in the media content's metadata can be used to identify whether the media content is AIGC.

[0185] The following example uses media content as an image and JUMBF (JPEG Universal MetadataBox Format) boxes as metadata to further explain the metadata generation process.

[0186] First, let's introduce the JUMBF frame.

[0187] Figure 3a This is a schematic diagram of the structure of a JUMBF frame as an example.

[0188] Reference Figure 3a For example, a JUMBF box is a type of super box. A JUMBF super box must include a JUMBF description box, n (n is a positive integer) JUMBF content boxes, and other boxes (e.g., padding boxes, or other types of boxes). The JUMBF description box precedes the JUMBF content boxes.

[0189] For example, a JUMBF Description Box can be used to describe a JUMBF Content Box; a JUMBF Content Box can include information related to the media content; and a Padding Box can be used for data completion. Of course, other types of Boxes can also be defined, and this application does not limit this.

[0190] Figure 3b This is a schematic diagram of the structure of the JUMBF Description Box as an example.

[0191] Reference Figure 3b For example, the JUMBF Description Box may include multiple fields: TYPE field (type field), TOGGLES(T) field (enabled status field), LABEL field (content request field), ID field (identity field), and SHA256HASH field (field used to store the sixth hash value).

[0192] It should be understood that, Figure 3b This is merely one example of the present application; the present application may include more than Figure 3b This application makes no limitation on the number of fields, whether more or fewer.

[0193] For example, the TYPE field (which can be called the JUMBF type field or type identifier) ​​can be used to describe the JUMBF content type (that is, the type of the JUMBF content box).

[0194] For example, the TYPE field can contain a 16-byte UUID whose value specifies the type of the JUMBFContent Box.

[0195] For example, the types of JUMBF Content Boxes can include, but are not limited to: Codestream ContentType, UUID Content Type, XML Content Type, JSON Content Type, CBOR Content Type, Embedded File Content Type, and other content types (which can be customized).

[0196] For example, the TYPE field is a required field.

[0197] For example, the TOGGLES(T) field (also known as the enable status flag) can contain a 1-byte (i.e., 8-bit) binary number that can be used to represent the value of the option associated with the JUMBF box. Each bit represents an option; when the value of the bit is 1, it indicates that the option is enabled; when the value of the bit is 0, it indicates that the option is disabled.

[0198] For example, the LABEL field (also known as a content request identifier) ​​can contain a variable-length text label that can be used to reference or request the content of the JUMBF box.

[0199] For example, the ID field (also known as the identity identifier) ​​can contain a unique 4-byte ID assigned by the user, which can be used to reference the JUMBF box in binary.

[0200] For example, the SHA256HASH field can be used to store a sixth hash value, which is determined based on n JUMBFContentBoxes. For example, the sixth hash value can be obtained by concatenating the n JUMBFContentBoxes in their order within the JUMBF SuperBox and then performing a hash calculation.

[0201] For example, the TOGGLES(T) field, LABEL field, ID field, and SHA256HASH field are optional fields.

[0202] based on Figure 3a and Figure 3b In one possible approach, this application can define a new content type (ContentType) in JUMBF, which can be called AIGCContentType, and the corresponding JUMBFContentBox can be called AICGContentBox. Furthermore, for the TYPE field in JUMBFDescriptionBox, a new identifier value (which can be called a preset identifier value) can be defined. The preset identifier value can indicate that JUMBFContentBox is AICGContentBox (or indicate that the type of JUMBFContentBox is AIGCContentType, which can also be understood as indicating that the media content is AIGC). In this case, Figure 2 The AIGC identifier in the JUMBF DescriptionBox can refer to the TYPE field.

[0203] Figure 3c This is a schematic diagram of the structure of AICGContentBox as an example.

[0204] Reference Figure 3c For example, AICGContentBox may include, but is not limited to: AIGCI field, AIModel (AI model) field, AIModelVer (AI Model Version) field, ParaSize (ParameterSize, AI model parameter data size) field, DataSize (AI model training set data size) field, and Data (load data) field.

[0205] For example, the AIGCI field (also known as the AIGC identifier, where the AIGC identifier represented by the TYPE field in JUMBFDescriptionBox and the AIGC identifier represented by the AIGCI field in AICGContentBox are two different identifiers, but they serve the same purpose) can also indicate whether JUMBFContentBox is an AICGContentBox. For instance, two identifier values ​​can be defined for the AIGCI field: "1" and "0"; "1" indicates that JUMBFContentBox is an AICGContentBox, and "0" indicates that JUMBFContentBox is not an AICGContentBox.

[0206] For example, the AIModel field (also known as the model identifier) ​​can indicate the type of AI model that generates the media content, such as ChatGPT (an artificial intelligence-driven natural language processing tool), Midjourney (an AI drawing tool), etc.

[0207] For example, the AIModelVer field (also known as the model version identifier) ​​can indicate the version of the AI ​​model that generates the media content; such as MidjourneyV4, MidjourneyV5, ChatGPTv3.5, ChatGPTv4, etc.

[0208] For example, the ParaSize field (also known as the model parameter data volume identifier) ​​can indicate the size of the AI ​​model parameters used to generate media content, such as 1 million parameters or 10 million parameters.

[0209] For example, the DataSize field (also known as the model training set data size identifier) ​​can indicate the size of the training set data of the AI ​​model used to generate media content, such as 1G or 47T.

[0210] For example, the Data field can be used to store information related to media content.

[0211] It should be noted that, in Figure 3c In the AICGContentBox, the Data field is required, while the other fields are optional.

[0212] Figure 4a This is a schematic diagram illustrating the metadata generation process as an example. Figure 4a In order to be in Figures 3a to 3c This is based on the above.

[0213] S401, acquire image.

[0214] S402, determine whether the image is AIGC.

[0215] For example, when an image is determined to be AIGC, S403 can be executed.

[0216] S403, generate AICGContentBox.

[0217] Assuming AICGContentBox is like Figure 3cAs shown, the value of the AIGCI field can be set to 1. When the image is AIGC, the image information can also include information about the AI ​​model that generated the image, such as the AI ​​model type, AI model version number, the magnitude of the model parameters, and the size of the training set. Then, based on the AI ​​model type, AI model version number, the magnitude of the model parameters, and the size of the training set, the AIModel, AIModelVer, ParaSize, and DataSize fields can be filled sequentially. The Data field can also be filled with image-related information. This results in the AICGContentBox.

[0218] For example, AICGContentBox can be a single object; in this case, the Data field of the AICGContentBox can contain all the relevant information about the image.

[0219] For example, there can be multiple AICGContentBoxes; in this case, the Data field of each AICGContentBox can contain some information from the image's related information. The image related information contained in the Data field of each AICGContentBox is different.

[0220] It should be understood that this application does not limit the number of AICGContentBoxes generated during the media asset generation phase.

[0221] S404, generate JUMBFDescriptionBox and set the value of the TYPE field in JUMBFDescriptionBox to the preset identifier value.

[0222] For example, the preset identifier value can be an identifier value that indicates JUMBFContentBox as AICGContentBox, and can be a 16-byte UUID.

[0223] For example, you can set the value of the TYPE field in JUMBFDescriptionBox to a preset identifier value and fill in other fields to obtain JUMBFDescriptionBox.

[0224] S405 encapsulates AICGContentBox and JUMBFDescriptionBox to obtain the JUMBF box.

[0225] For example, you can encapsulate AICGContentBox and JUMBFDescriptionBox to get JUMBFSuperBox, which is a JUMBF box; thus, you get the image's metadata.

[0226] It should be understood that prior to S405, OtherBox could also be generated. In this case, S405 could encapsulate AICGContentBox, JUMBFDescriptionBox, and OtherBox to obtain a JUMBF box. OtherBox can be PaddingBox or other types of Box defined as the standard evolves; this application does not impose any restrictions on this.

[0227] Figure 4b This is a schematic diagram of the structure of a JUMBF frame as an example.

[0228] Reference Figure 4b For example, the JUMBF bounding box of an AI-generated image may include: JUMBFDescriptionBox, n AICGContentBoxes (which may include AICGContentBox1, AICGContentBox2, ..., AICGContentBoxn) and OtherBox; wherein, "TYPE: AIGC" in JUMBFDescriptionBox indicates that the ContentType of JUMBFContentBox is AIGCContentType.

[0229] It should be noted that both the JUMBFDescriptionBox and AICGContentBox in the JUMBF box have identifiers (jumd); this allows the JUMBF Description Box and AICG ContentBox to be read later based on the identifiers.

[0230] S406, based on the image and the JUMBF frame, obtain the JPEG file.

[0231] For example, a JPEG file can be created first, which includes multiple marker segments. Then, an image and a JUMBF frame can be embedded into different marker segments of the JPEG file. For instance, a JUMBF frame can be embedded into the APP11 marker segment of the JPEG file.

[0232] It should be understood that in S403, when it is determined that the image is not AIGC, a JUMBF Content Box of other content types can be generated, as well as a JUMBF Description Box, and the field value of the TYPE field in the JUMBF Description Box can be set to other identifier values ​​(indicating content types other than AIGC Content Type); then, S405 to S406 are executed.

[0233] based on Figure 3a and Figure 3b In one possible approach, this application could add a field to the JUMBF DescriptionBox to indicate whether the media content is AIGC.

[0234] Figure 5a This is a schematic diagram of the structure of the JUMBF Description Box as an example.

[0235] Reference Figure 5a For example, in Figure 3b Building upon this, a new field, AIGCI (AIGC Identify), has been added to the JUMBF Description Box. The AIGCI field indicates whether the media content is AIGC. The AIGCI field in the JUMBF Description Box can be used as... Figure 2 The AIGC identifier in the embodiment. For example, multiple identifier values ​​can be defined for the AIGCI field (including a preset identifier value and at least one other identifier value); for example, two identifier values ​​"1" and "0" can be defined. "1" can be the preset identifier value, indicating that the media content is AIGC, and "0" can be another identifier value, indicating that the media content is not AIGC.

[0236] Figure 5b This is a schematic diagram illustrating the metadata generation process as an example. Wherein, Figure 5b Is Figure 5a This is based on the above.

[0237] S501, acquire image.

[0238] S502, determine whether the image is AIGC.

[0239] S503 generates the JUMBF Content Box.

[0240] For example, the relevant information of the image can be encapsulated into a JUMBF Content Box corresponding to any of the following content types: Codestream Content Type, UUID Content Type, XML Content Type, JSON Content Type, CBOR Content Type, and EmbeddedFile Content Type.

[0241] Figure 5c This is a schematic diagram of the structure of the JUMBF Content Box as an example.

[0242] Reference Figure 5c (1) For example, if the Content Type of the JUMBF Content Box is CodestreamContentType, then the JUMBF Content Box can be called a Codestream Box. The Codestream Box includes a Code field, which stores information related to the media content. For example, the relevant information of the image can be encoded and filled into the Code field to obtain the Codestream Box.

[0243] Reference Figure 5c (2) For example, the Content Type of the JUMBF Content Box is XML ContentType or JSON ContentType; in this case, the JUMBF Content Box can be called an XML Box or a JSON Box. The XML Box or JSON Box can include a Data field, which is used to store relevant information about the media content. For example, the relevant information about the image can be represented using XLM and then filled into the Data field to obtain an XML Box. For example, the relevant information about the image can be represented using JSON and then filled into the Data field to obtain a JSON Box.

[0244] Reference Figure 5c(3) For example, the ContentType of JUMBFContentBox is UUIDContentType; in this case, JUMBFContentBox can be called UUIDBox. UUIDBox can include an ID field and a Data field; the ID field can include a 16-byte UUID, indicating the interpretation of specific information stored in the Data field, which is used to store information related to the media content. For example, the image's relevant information can be represented by a UUID and filled into the Data field, and the ID field can be filled accordingly to obtain a UUIDBox.

[0245] For example, a JUMBFContentBox can be a single JUMBFContentBox; in this case, the Data field in the JUMBFContentBox can contain all the relevant information about the image.

[0246] For example, there can be multiple JUMBFContentBoxes; in this case, the Data field in each JUMBFContentBox can contain some information from the image's related information; the image-related information contained in the Data field of each JUMBFContentBox is different.

[0247] S504, generate JUMBFDescriptionBox and set the value of the AIGCI field in the DescriptionBox to the preset identifier value.

[0248] For example, during the generation of JUMBFDescriptionBox, the value of the AIGCI field in JUMBFDescriptionBox can be set to a preset identifier value, and other fields can be filled to obtain JUMBFDescriptionBox.

[0249] It should be noted that, in Figure 5b In one embodiment, for the TYPE field, the identifier value of the TYPE field can be set to other identifier values ​​(indicating a content type other than AIGCContentType).

[0250] S505, Based on the JUMBF content box and the JUMBF description box, obtain the JUMBF box.

[0251] S506, based on the image and the JUMBF frame, obtain the JPEG file.

[0252] For example, S505 to S506 can be referred to the description of S405 to S406 above, and will not be repeated here.

[0253] Figure 5d This is a schematic diagram of the structure of a JUMBF frame as an example.

[0254] Reference Figure 5d For example, the JUMBF box of an AI-generated image may include: JUMBFDescriptionBox, n XMLBoxes (which may include XMLBox1, XMLBox2, ..., XMLBoxn) and OtherBox; where "AIGCI: 1" in JUMBFDescriptionBox indicates that the image is AIGC, and "TYPE: XML" in JUMBFDescriptionBox indicates that the ContentType of JUMBFContentBox is XMLContentType.

[0255] based on Figure 3a and Figure 3b In one possible approach, this application may add a field to the JUMBFContentBox to indicate whether the media content is AIGC.

[0256] Figure 6a This is a schematic diagram of the structure of JUMBFContentBox as an example.

[0257] Reference Figure 6a (1), for example, in Figure 5c Based on (1), add a new field, AIGCI, to CodestreamBox. The AIGCI field indicates whether the media content is AIGC. The AIGCI field in CodestreamBox can be used as... Figure 2 AIGC identifier in the embodiment.

[0258] Reference Figure 6a (2), for example, in Figure 5c Based on (2), add a new field, AIGCI, to the XMLBox or JSONBox. The AIGCI field indicates whether the media content is AIGC. The AIGCI field in the XMLBox or JSONBox can be used as... Figure 2 AIGC identifier in the embodiment.

[0259] Reference Figure 6a (3), for example, in Figure 5c Based on (3), a new field AIGCI is added to the UUIDBox. The AIGCI field indicates whether the media content is AIGC. The AIGCI field in the UUIDBox can be used as... Figure 2 AIGC identifier in the embodiment.

[0260] For example, multiple identifier values ​​can be defined for the AIGCI field (including a preset identifier value and at least one other identifier value). For instance, two identifier values ​​can be defined: "1" and "0"; "1" can be the preset identifier value, indicating that the media content is AIGC, and "0" indicates that the media content is not AIGC.

[0261] Figure 6b This is a schematic diagram illustrating the metadata generation process as an example. Wherein, Figure 6b Is Figure 6a This is based on the above.

[0262] S601, acquire image.

[0263] S602, determine whether the image is AIGC.

[0264] S603, generate JUMBFContentBox and set the value of the AIGCI field in JUMBFContentBox to the preset identifier value.

[0265] For example, such as Figure 6a As shown in (1), the relevant information of the image can be encoded and filled into the Code field, and the field value of the AIGCI field can be set to the preset identifier value to obtain the CodestreamBox.

[0266] For example, such as Figure 6a As shown in (2), the image information can be represented in XML and then filled into the Data field, and the value of the AIGCI field can be set to a preset identifier value to obtain an XMLBox. Alternatively, the image information can be represented in JSON and then filled into the Data field, and the value of the AIGCI field can be set to a preset identifier value to obtain a JSONBox.

[0267] For example, such as Figure 6a As shown in (3), the relevant information of the image can be represented by UUID and filled into the Data field, and the field value of the AIGCI field can be set to the preset identifier value to obtain the UUIDBox.

[0268] S604, generate JUMBFDescriptionBox.

[0269] For example, during the generation of JUMBFDescriptionBox, the value of the TYPE field in JUMBFDescriptionBox can be set to other identifier values, and other fields can be filled to obtain JUMBFDescriptionBox.

[0270] S605, Based on the JUMBF content box and the JUMBF description box, obtain the JUMBF box.

[0271] S606, based on the image and the JUMBF frame, obtains the JPEG file.

[0272] For example, S605 to S606 can be referred to the description of S405 to S406 above, and will not be repeated here.

[0273] Figure 6c This is a schematic diagram of the structure of a JUMBF frame as an example.

[0274] Reference Figure 6c For example, the JUMBF bounding box of an AI-generated image may include: JUMBFDescriptionBox, XMLBox (which may include XMLBox1, XMLBox2, ..., XMLBoxn) and OtherBox; wherein, "TYPE: XML" in JUMBFDescriptionBox indicates that the ContentType of JUMBFContentBox is XMLContentType, and "AIGCI: 1" in XMLBox indicates that the image is AIGC.

[0275] For example, the metadata of media content may also include a DNT (DonotTrain) field (also known as a training enable flag), which can indicate whether the media content is allowed to be used for model training.

[0276] For example, two flag values ​​can be defined for the DNT field, such as "1" and "0". "1" can indicate that the media content is allowed to be used for model training, and "0" indicates that the media content is not allowed to be used for model training.

[0277] Figure 7a This is a schematic diagram illustrating the structure of the JUMBFDescriptionBox as an example. Figure 7a In order to be in Figure 3b This is based on the above.

[0278] Reference Figure 7a For example, JUMBFDescriptionBox may also include a DNT field.

[0279] Figure 7b This is a schematic diagram illustrating the structure of an AIGCContentBox as an example. Figure 7b In order to be in Figure 3c This is based on the above.

[0280] Reference Figure 7bFor example, AIGCContentBox may also include a DNT field.

[0281] It should be understood that the DNT field is an optional field, and this application does not restrict whether the DNT field is located in the JUMBFContentBox or the JUMBFDescriptionBox.

[0282] Figure 7c This is a schematic diagram illustrating the structure of the JUMBFDescriptionBox as an example. Figure 7c In order to be in Figure 3b This is based on the above.

[0283] For example, the metadata of media content may also include a first hash value. (See reference...) Figure 7c For example, a new field called HASH can be added to JUMBFDescriptionBox, with the first hash value located in this HASH field. It should be understood that the HASH field is optional, and the first hash value is optional information.

[0284] For example, when JUMBFDescriptionBox only includes the AIGC identifier, a hash algorithm can be used to calculate the identifier value of the AIGC identifier to obtain the first hash value.

[0285] For example, when the JUMBFDescriptionBox includes an AIGC identifier and at least one of the following fields: TOGGLES(T), LABEL, ID, SHA256HASH, and DNT, a hash algorithm can be used to calculate the identifier value of the concatenated AIGC identifier and the field values ​​of at least one of the TOGGLES(T), LABEL, ID, SHA256HASH, and DNT fields to obtain a first hash value. It should be understood that this application does not limit the hash algorithm used to calculate the first hash value. This prevents the field values ​​in the JUMBFDescriptionBox from being tampered with.

[0286] Figure 7d This is a schematic diagram illustrating the structure of the JUMBFDescriptionBox as an example. Figure 7d In order to be in Figure 7c This is based on the above.

[0287] For example, the metadata of media content may also include a digital signature of the first hash value. (See reference...) Figure 7dFor example, a new field SIGNATURE can be added to JUMBFDescriptionBox, where the digital signature of the first hash value is located. Specifically, the first hash value can be digitally signed using a private key based on a digital signature algorithm. It should be understood that this application does not restrict the digital signature algorithm used to calculate the digital signature of the first hash value. This prevents the fields in JUMBFDescriptionBox from being forged.

[0288] It should be understood that the SIGNATURE field is an optional field, and the digital signature of the first hash value is optional information.

[0289] Figure 7e This is a schematic diagram illustrating the structure of the Data field in a JUMBFContentBox as an example.

[0290] For example, the metadata of media content may also include: a second hash value, a third hash value, and a digital signature of the third hash value. (See reference...) Figure 7e (1) For example, the digital signatures of the second hash value, the third hash value, and the third hash value can be located in the Data field of the JUMBFContentBox.

[0291] For example, the second hash value is the hash value of the media content. Specifically, a hash algorithm can be used to hash the media content to obtain the second hash value. It should be understood that this application does not limit the hash algorithm used to calculate the second hash value.

[0292] For example, the third hash value is determined based on the relevant information of the media content and the second hash value. Specifically, the relevant information of the media content and the second hash value can be combined (also called concatenated or cascaded); a hash algorithm is used to perform hash calculation on the combined relevant information of the media content and the second hash value to obtain the third hash value. It should be understood that this application does not limit the hash algorithm used to calculate the third hash value.

[0293] For example, a digital signature of the third hash value can be obtained by digitally signing the third hash value using a private key based on a digital signature algorithm. It should be understood that this application does not limit the digital signature algorithm used to calculate the third hash value.

[0294] For example, the metadata of media content may also include: one or more media claims and one or more fourth hash values. (See also...) Figure 7e (2) For example, m (m is a positive integer) media claims and m fourth hash values ​​can be located in the Data field of JUMBFContentBox.

[0295] For example, a media statement is a statement about media content. The media statement may include copyright information, thumbnails, whether it can be modified, operating permissions, usage permissions, the source of the media content, etc. This application does not limit this.

[0296] For example, the fourth hash value is the hash value of the media claim, thus preventing the media claim from being tampered with. Specifically, a hash algorithm can be used to hash the media claim to obtain the fourth hash value; where one media claim corresponds to one fourth hash value. It should be understood that this application does not limit the hash algorithm used to calculate the fourth hash value.

[0297] For example, the metadata of media content may also include a fifth hash value; see reference Figure 7e (3) For example, the fifth hash value can be located in the Data field of JUMBFContentBox.

[0298] For example, the fifth hash value can be determined based on the second hash value, the third hash value, the digital signature of the third hash value, and one or more fourth hash values.

[0299] For example, the intermediate hash value can be obtained by concatenating the second hash value, the third hash value, the digital signature of the third hash value, and one or more of the fourth hash values.

[0300] If the intermediate hash value is obtained by hashing the second hash value, the third hash value, the digital signature of the third hash value, and all items in the fourth hash value, then the intermediate hash value can be used as the fifth hash value.

[0301] When the intermediate hash value is obtained by concatenating the second hash value, the third hash value, the digital signature of the third hash value, and a portion of one or more fourth hash values, the fifth hash value can be obtained by performing a hash calculation based on the intermediate hash value, the second hash value, the third hash value, the digital signature of the third hash value, and another portion of one or more fourth hash values.

[0302] For example, a hash tree can be used to perform hash calculations on the second hash value, the third hash value, the digital signature of the third hash value, and one or more fourth hash values ​​to obtain the fifth hash value. Here, a hash tree is a data structure that can include, but is not limited to, binary trees, ternary trees, Merkle trees, etc., and this application does not impose any restrictions on it.

[0303] In one possible approach, the fifth hash value is the value of the root node of the hash tree, and the second, third, and digital signatures of the third hash value and one or more fourth hash values ​​are the values ​​of the leaf nodes of the hash tree, respectively.

[0304] Figure 7f This is a schematic diagram illustrating the fifth hash value generation process as an example.

[0305] exist Figure 7f In this example, the hash tree is a Merkle Tree. The fifth hash value is the value of the root node (MerkleTreeRoot, MTR) of the Merkle Tree, which can be represented by MTR. The second hash value (MediaHash), the third hash value (Hash3), the digital signature of the third hash value (Signature1), and the two fourth hash values ​​(ClaimHashN1 and ClaimHashN2) are the values ​​of the leaf nodes of the Merkle Tree.

[0306] Reference Figure 7f For example, hash calculations can be performed on the concatenated MediaHash and Hash3 to obtain HashX1 (the intermediate hash value); hash calculations can be performed on the concatenated Signature1 and ClaimHashN1 to obtain HashX2 (the intermediate hash value); and hash calculations can be performed on ClaimHashN2 to obtain HashX3 (the intermediate hash value). Then, hash calculations can be performed on the concatenated HashX1 and HashX2 to obtain HashX4; and hash calculations can be performed on HashX3 to obtain HashX5. Finally, hash calculations can be performed on the concatenated HashX4 and HashX5 to obtain the MTR.

[0307] It should be understood that, Figure 7f This is just one example of a MerkleTree. The MerkleTree in this application can be other data structures, and this application does not limit it.

[0308] It should be understood that this application does not restrict which leaf node of the MerkleTree corresponds to the second hash value, the third hash value, the digital signature of the third hash value, and one or more fourth hash values.

[0309] In one possible approach, the fifth hash value is the value of the root node of the hash tree, the third hash value is the value of the root node of the left subtree of the hash tree, and the digital signatures of the second and third hash values ​​and one or more fourth hash values ​​are the values ​​of the leaf nodes of the right subtree of the hash tree, respectively.

[0310] Figure 7g This is a schematic diagram illustrating the fifth hash value generation process as an example.

[0311] exist Figure 7g In this example, the hash tree is a Merkle Tree. The fifth hash value is the value of the root node of the Merkle Tree, which can be represented using MTR. The third hash value (Hash3) is the value of the root node of the left subtree of the Merkle Tree. The second hash value (MediaHash), the digital signature of the third hash value (Signature1), and the two fourth hash values ​​(ClaimHash N1 and ClaimHashN2) are the values ​​of the leaf nodes of the right subtree of the Merkle Tree, respectively.

[0312] Reference Figure 7g For example, the concatenated MediaHash and Signature1 can be hashed to obtain HASHX6 (the intermediate hash value), and the concatenated ClaimHashN1 and ClaimHashN2 can be hashed to obtain HashX7 (the intermediate hash value). Then, the concatenated HashX6 and HashX7 are hashed to obtain HashX8. Finally, the concatenated HashX8 and Hash3 are hashed to obtain MTR.

[0313] It should be understood that this application does not restrict which leaf node of the right subtree of the MerkleTree corresponds to the digital signatures of the second hash value, the third hash value, and one or more fourth hash values.

[0314] Next, the process of identifying media content will be explained.

[0315] Figure 8 This is a schematic diagram illustrating the identification process as an example. Figure 8 The example is with Figure 2 The identification process corresponding to the metadata generation process in the embodiment.

[0316] S801, Obtain the metadata of the media content, which includes the AIGC identifier.

[0317] For example, as described in the above embodiments, the metadata of the media content includes an AIGC identifier that indicates whether the media content is AIGC; therefore, for any media content in an electronic device (which may be media content contained in media assets generated by the electronic device or media content contained in media assets received by the electronic device from other electronic devices), when it is necessary to identify whether the media content is AIGC, the metadata of the media content can be obtained; and then, based on the metadata of the media content, it can be identified whether the media content is AIGC.

[0318] S802, reads the AIGC identifier from the metadata of the media content.

[0319] S803 determines whether the media content is AIGC based on the AIGC identifier.

[0320] For example, the AIGC identifier can be determined from the metadata of the media content; then the value of the AIGC identifier can be read; and then, based on the value of the AIGC identifier, it can be determined whether the media content is AIGC.

[0321] For example, when the value of the AIGC identifier is a preset identifier value, the media content is determined to be AIGC. When the value of the AIGC identifier is another identifier value, the media content is determined not to be AIGC.

[0322] The following example uses media content as an image and JUMBF bounding boxes as metadata to further illustrate the process of media content recognition.

[0323] For example, in Figure 4a In this embodiment, the JUMBFDescriptionBox contains a TYPE field. When the TYPE field is an AIGC identifier, the process for identifying whether an image is AIGC can be referred to... Figure 9 Description of the embodiments.

[0324] Figure 9 This is a schematic diagram illustrating an exemplary recognition process. Figure 9 The embodiments are with Figure 4a The identification process corresponding to the metadata generation process in the embodiment.

[0325] S901, obtain the JUMBF bounding box of the image. The JUMBFDescriptionBox in the JUMBF bounding box includes the AIGC identifier.

[0326] For example, based on Figure 4a As described in the embodiment, the TYPE field in the JUMBFDescriptionBox of the image's JUMBF bounding box is an AIGC identifier; therefore, when it is necessary to determine whether an image is AIGC, the JUMBF bounding box of the image can be obtained.

[0327] For example, a JUMBF frame can be read from the APP11 marker segment of a JPEG file. Specifically, after reading the first identifier (jumd) in the APP11 marker segment of the JPEG file, data reading can continue until the second identifier (jumd) is read, confirming that the first JUMBF frame has been read. After reading the second identifier (jumd) in the APP11 marker segment of the JPEG file, data reading can continue until the third identifier (jumd) is read, confirming that the second JUMBF frame has been read, and so on.

[0328] S902, Read the field value of the TYPE field from the JUMBFDescriptionBox of the JUMBF box.

[0329] For example, after reading each JUMBF box, the data within that JUMBF box can be read. Specifically, after reading the first identifier (jumd) in the JUMBF box, data reading continues until the second identifier (jumd) is read. At this point, it can be determined that the JUMBFDescriptionBox in that JUMBF box has been read. Then, the value of the TYPE field in the JUMBFDescriptionBox is read to obtain the value of the AIGC identifier.

[0330] S903, determine whether the value of the TYPE field is a preset identifier value.

[0331] Then, it can be determined whether the value of the TYPE field is the preset identifier value, that is, whether the value of the AIGC identifier is the preset identifier value.

[0332] For example, when the value of the TYPE field is a preset identifier value, the image can be determined to be AIGC; at this time, reading the next JUMBF box in the APP11 marker segment can be stopped.

[0333] For example, when the value of the TYPE field is not the preset identifier value, the next JUMBF box can be read in the APP11 mark segment, and then S902 and S903 can be executed; when the value of the TYPE field included in the JUMBFDescriptionBox of all JUMBF boxes in the APP11 mark segment is not the preset identifier value, it can be determined that the image is not AIGC.

[0334] S904, when the value of the AIGC identifier is a preset identifier value, the image is determined to be AIGC.

[0335] S905, when the value of the AIGC identifier is not a preset identifier value, the image is determined to be not AIGC.

[0336] For example, in Figure 4a and Figure 6b In this embodiment, the JUMBFContentBox contains an AIGCI field. When the AIGCI field is an AIGC identifier, the process for identifying whether an image is AIGC can be referred to... Figure 10 Description of the embodiments.

[0337] Figure 10 This is a schematic diagram illustrating an exemplary recognition process. Figure 10 The embodiments are with 4a and Figure 6bThe identification process corresponding to the metadata generation process in the embodiment.

[0338] S1001, Obtain the JUMBF box of the image. The JUMBFContentBox in the JUMBF box includes the AIGCI field.

[0339] For example, based on Figure 4a and Figure 6b As described in the embodiments, when the JUMBF box of an image includes the AIGCI field, if it is necessary to determine whether an image is AIGC, the JUMBF box of that image can be obtained. The specific process of reading the JUMBF box of an image can be referred to the description of S901 above, and will not be repeated here.

[0340] S1002, Read the field value of the AIGCI field from the JUMBFContentBox of the JUMBF box.

[0341] For example, after each JUMBF box is read, the data within that JUMBF box can be read. Specifically, after the second identifier (jumd) is read in the JUMBF box, data reading continues until the third identifier (jumd) is read, at which point it can be determined that the first JUMBFContentBox in the JUMBF box has been read. After the second identifier (jumd) is read, data reading continues until the third identifier (jumd) is read, at which point it can be determined that the second JUMBFContentBox in the JUMBF box has been read, and so on.

[0342] After reading each JUMBFContentBox, the value of the AIGCI field in the AIGCI field of that JUMBFContentBox can be read; then S1003 is executed.

[0343] S1003, determine whether the value of the AIGCI field is the preset identifier value.

[0344] For example, when the value of the AIGCI field is a preset identifier value, the image can be determined to be AIGC; at this time, it is possible to stop reading the next JUMBFContentBox from the JUMBF box, and it is possible to stop reading the next JUMBF box in the APP11 marker segment.

[0345] For example, when the value of the AIGCI field is not a preset identifier value, the next JUMBF ContentBox can be read from the current JUMBF box, and then S1002-S1003 are executed. When the value of the AIGCI field in all JUMBF ContentBoxes within the current JUMBF box is not a preset identifier value, the next JUMBF box can be read from the APP11 marker segment. When the value of the AIGCI field in all JUMBF ContentBoxes within all JUMBF boxes in the APP11 marker segment is not a preset identifier value, it can be determined that the image is not AIGC.

[0346] S1004, when the value of the AIGCI field is a preset identifier value, the image is determined to be AIGC.

[0347] S1005, when the value of the AIGCI field is not a preset identifier value, determine that the image is not AIGC.

[0348] For example, in Figure 5b In this embodiment, the JUMBFDescriptionBox contains an AIGCI field. When the AIGCI field is an AIGC identifier, the process of identifying whether an image is AIGC can be referred to... Figure 11 Description of the embodiments.

[0349] Figure 11 This is a schematic diagram illustrating an exemplary recognition process. Figure 11 The embodiments are with Figure 5b The identification process corresponding to the metadata generation process in the embodiment.

[0350] S1101, Obtain the JUMBF bounding box of the image. The JUMBFDescriptionBox of the JUMBF bounding box includes the AIGC identifier.

[0351] S1102, Read the field value of the AIGC field from the JUMBFDescriptionBox of the JUMBF box.

[0352] For example, S1001 to S1002 can be referred to the description of S801 to S802 above, and will not be repeated here.

[0353] It should be noted that after reading the JUMBFDescriptionBox in S1102, the value of the AIGCI field in the JUMBFDescriptionBox can be read to obtain the value of the AIGC identifier.

[0354] S1103, Determine whether the value of the AIGC field is a preset identifier value.

[0355] S1104, when the value of the AIGC field is a preset identifier value, the image is determined to be AIGC.

[0356] S1105, when the value of the AIGC field is not a preset identifier value, determine that the image is not AIGC.

[0357] Furthermore, when media content is needed for AI model training, the DNT field value can be read from the media content's metadata. Then, based on the DNT field value, it can be determined whether the media content is permitted for AI model training. If the DNT field value indicates that the media content is permitted, it can be used as a training sample in the AI ​​model's training set. If the DNT field value indicates that the media content is not permitted, it is unnecessary to add it to the AI ​​model's training set. For details on how to read the DNT field value, please refer to the above. Figures 8-11 The description of how the AIGC identifier value is read in the embodiment will not be repeated here.

[0358] Figure 12 This is a schematic diagram of an exemplary metadata generation apparatus. This metadata generation apparatus can be used to execute the methods of the foregoing embodiments; therefore, the beneficial effects it achieves can be referred to in the corresponding methods provided above, and will not be repeated here.

[0359] Reference Figure 12 For example, the metadata generation apparatus may include:

[0360] Content acquisition module 1201 is used to acquire media content;

[0361] Metadata generation module 1202 is used to generate metadata for media content. The metadata for media content includes an AIGC identifier for AI content, which indicates whether the media content is AIGC.

[0362] For example, the metadata generation module 1202 is also used to obtain the media assets corresponding to the media content based on the media content and the metadata of the media content.

[0363] Figure 13 This is a schematic diagram of an exemplary identification device. The metadata generation device can be used to execute the methods of the foregoing embodiments; therefore, the beneficial effects it achieves can be referred to in the corresponding methods provided above, and will not be repeated here.

[0364] Reference Figure 13 For example, the identification device may include:

[0365] Metadata acquisition module 1301 is used to acquire metadata of media content;

[0366] The reading module 1302 is used to read the AIGC identifier from the metadata of the media content;

[0367] The judgment module 1303 is used to determine whether the media content is AIGC based on the AIGC identifier.

[0368] In one example, Figure 14 A schematic block diagram illustrating an embodiment of the present application shows an apparatus 1400. The apparatus 1400 may include a processor 1401 and a transceiver / transceiver pin 1402, and optionally, a memory 1403.

[0369] The various components of device 1400 are coupled together via bus 1404, which includes a data bus, a power bus, a control bus, and a status signal bus. However, for clarity, all buses are referred to as bus 1404 in the figure.

[0370] Optionally, the memory 1403 can be used to store instructions from the foregoing method embodiments. The processor 1401 can be used to execute the instructions in the memory 1403, control the receive pin to receive signals, and control the transmit pin to transmit signals.

[0371] Device 1400 may be an electronic device or a chip of an electronic device in the above method embodiments.

[0372] All relevant content of each step involved in the above method embodiments can be referenced from the functional description of the corresponding functional module, and will not be repeated here.

[0373] This application also provides a chip, including one or more interface circuits and one or more processors; the one or more processors receive or send data through the one or more interface circuits, and when the one or more processors execute computer instructions, the steps of the above embodiments are performed. The interface circuit is a transceiver / transceiver pin 1402.

[0374] This embodiment also provides a computer-readable storage medium storing computer instructions. When the computer instructions are executed on an electronic device, the electronic device performs the aforementioned method steps to implement the methods described in the above embodiments.

[0375] This embodiment also provides a computer program product containing computer instructions that, when executed by a computer or processor, cause the computer to perform the aforementioned related steps to implement the methods described in the above embodiments.

[0376] In addition, embodiments of this application also provide an apparatus, which may specifically be a chip, component, or module. The apparatus may include a connected processor and a memory; wherein the memory is used to store computer execution instructions, and when the apparatus is running, the processor may execute the computer execution instructions stored in the memory to cause the chip to execute the methods in the above-described method embodiments.

[0377] In this embodiment, the electronic device, computer-readable storage medium, computer program product or chip are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can be referred to the beneficial effects of the corresponding methods provided above, and will not be repeated here.

[0378] Through the above description of the embodiments, those skilled in the art will understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0379] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another apparatus, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0380] The units described as separate components may or may not be physically separate. A component shown as a unit can be one or more physical units; that is, it can be located in one place or distributed in multiple different locations. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0381] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0382] Any content in the various embodiments of this application, as well as any content in the same embodiment, can be freely combined. Any combination of the above content is within the scope of this application.

[0383] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0384] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0385] The steps of the methods or algorithms described in conjunction with the embodiments of this application can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium well known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.

[0386] Those skilled in the art will recognize that the functions described in the embodiments of this application in one or more of the above examples can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer-readable storage media and communication media, wherein communication media include any medium that facilitates the transmission of a computer program from one place to another. Storage media can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0387] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for generating metadata, characterized in that, The method includes: Obtain media content, wherein the media content is media content that does not contain metadata; Generate metadata for the media content, the metadata of the media content including an AIGC identifier for AIGC content, the AIGC identifier indicating whether the media content is AIGC; The metadata of the media content is a JPEG Universal Metadata Frame Format (JUMBF) frame, which includes a JUMBF description frame and a JUMBF content frame. The JUMBF description frame and / or the JUMBF content frame includes the AIGC identifier. The JUMBF content box also includes relevant information about the media content.

2. The method according to claim 1, characterized in that, The metadata of the media content also includes a first hash value, which is a hash value of the data including the AIGC identifier.

3. The method according to claim 2, characterized in that, The metadata of the media content also includes a digital signature of the first hash value.

4. The method according to any one of claims 1 to 3, characterized in that, The metadata of the media content also includes a training enable flag, which indicates whether the media content is permitted for model training.

5. The method according to any one of claims 1 to 3, characterized in that, The metadata of the media content also includes at least one of the following: model identifier, model version identifier, model parameter data volume identifier, or model training set data volume identifier.

6. The method according to any one of claims 1 to 3, characterized in that, The metadata of the media content also includes relevant information about the media content.

7. The method according to claim 6, characterized in that, The metadata of the media content also includes: a second hash value, a third hash value, and a digital signature of the third hash value, wherein the second hash value is the hash value of the media content, and the third hash value is a hash value that includes relevant information of the media content and data of the second hash value.

8. The method according to claim 7, characterized in that, The metadata of the media content also includes one or more media claims and one or more fourth hash values; the media claim is a claim on the media content, and the fourth hash value is the hash value of the media claim.

9. The method according to claim 8, characterized in that, The metadata of the media content also includes a fifth hash value, which is a hash value of data including the second hash value, the third hash value, the digital signature of the third hash value, and one or more fourth hash values.

10. The method according to any one of claims 1 to 3, 7 to 9, characterized in that, The method further includes: Based on the media content and its metadata, the media assets corresponding to the media content are obtained.

11. The method according to claim 1, characterized in that, The JUMBF content box also includes at least one of the following: model identifier, model version identifier, model parameter data volume identifier, or model training set data volume identifier.

12. The method according to claim 1 or 11, characterized in that, The relevant information of the media content includes at least one of the following: the generation time of the media content, the author's name of the media content, the digital content identifier of the media content, the generation location of the media content, the information of the media content generation device, the resolution of the media content, or the size of the media content.

13. The method according to claim 12, characterized in that, The JUMBF descriptor box also includes a first hash value, an enabled status identifier, a content request identifier, an identity identifier, and a sixth hash value; The sixth hash value is determined based on the JUMBF content frame, and the first hash value is a hash value of data including the AIGC identifier, the enabled status identifier, the content request identifier, the identity identifier, and the sixth hash value.

14. A method for identification, characterized in that, The method includes: Obtain metadata of media content; The metadata of the media content is a JPEG Universal Metadata Frame Format (JUMBF) frame, which includes a JUMBF description frame and a JUMBF content frame. The JUMBF description frame and / or the JUMBF content frame includes an AIGC identifier. The JUMBF content box also includes relevant information about the media content; Read the AIGC (Artificial Intelligence Generated Content) identifier from the metadata of the media content; Based on the AIGC identifier, determine whether the media content is AIGC.

15. The method according to claim 14, characterized in that, The method further includes: When the value of the AIGC identifier is a preset identifier value, the media content is determined to be AIGC.

16. A metadata generation apparatus, characterized in that, The device includes: The content acquisition module is used to acquire media content, wherein the media content is media content that does not have metadata; Metadata generation module, used to generate metadata of the media content, the metadata of the media content including AIGC identifier, the AIGC identifier indicating whether the media content is AIGC; The metadata of the media content is a JPEG Universal Metadata Frame Format (JUMBF) frame, which includes a JUMBF description frame and a JUMBF content frame. The JUMBF description frame and / or the JUMBF content frame includes the AIGC identifier. The JUMBF content box also includes relevant information about the media content.

17. An identification device, characterized in that, The device includes: The metadata acquisition module is used to acquire the metadata of the media content. The metadata of the media content is a JPEG Universal Metadata Frame (JUMBF) frame. The JUMBF frame includes a JUMBF description frame and a JUMBF content frame. The JUMBF description frame and / or the JUMBF content frame includes an AIGC identifier. The JUMBF content box also includes relevant information about the media content; The reading module is used to read the AIGC identifier of the artificial intelligence-generated content from the metadata of the media content; The determination module is used to determine whether the media content is AIGC based on the AIGC identifier.

18. An electronic device, characterized in that, include: A memory and a processor, wherein the memory is coupled to the processor; The memory stores program instructions that, when executed by the processor, cause the electronic device to perform the metadata generation method as described in any one of claims 1 to 13.

19. An electronic device, characterized in that, include: A memory and a processor, wherein the memory is coupled to the processor; The memory stores program instructions that, when executed by the processor, cause the electronic device to perform the identification method as described in claim 14 or claim 15.

20. A chip, characterized in that, It includes one or more interface circuits and one or more processors; the one or more processors receive or send data through the one or more interface circuits, and when the one or more processors execute computer instructions, the steps of the method as described in any one of claims 1 to 15 are performed.

21. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed on a computer or processor, causes the computer or processor to perform the method as described in any one of claims 1 to 15.

22. A computer program product, characterized in that, The computer program product includes computer instructions that, when executed by a computer or processor, cause the steps of the method as described in any one of claims 1 to 15 to be performed.

Images