Data processing method and electronic equipment

By collecting and processing the image and voice data of the target object in online video calls, generating confidence and prompt strategies, the authenticity challenges of video content caused by AI technology are solved, and the user's evaluation and experience of the target object's identity is improved. It is suitable for Various video call software.

CN118941442APending Publication Date: 2024-11-12LENOVO (BEIJING) LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411195549.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

In online video calls, the widespread application of AI technology has led to the challenge of the authenticity of video content, and it is difficult to effectively identify and verify the identity of the target object and the authenticity of the operating device.

Method used

By collecting image and voice data of the target object on the first device, processing using the target model, generating confidence indicators, and generating prompt strategies based on these indicators, limiting the degree of image processing, ensuring the authenticity of the output data and user experience.

Benefits of technology

It improves the authenticity assessment of the target object identity by users, enhances the user experience, and ensures the accuracy and rationality of the data through the combination of confidence and prompting strategies, and is suitable for different types of video call software.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941442A_ABST
    Figure CN118941442A_ABST
Patent Text Reader

Abstract

The invention discloses a data processing method and electronic equipment. The method comprises the steps that first equipment obtains target data; the target data is acquired based on a first acquisition module of the first equipment and at least comprises a first image of the target object, and the first image is obtained by processing a second image acquired by the first acquisition module through an image processing module; processing the target data based on the target model to obtain a target processing result; the target model runs on the first equipment, and the target processing result comprises parameters indicating the identity of the target object; and sending first information to the second device based on the target processing result and the target data, so that the second device outputs second information based on the first information.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a data processing method and an electronic device. Background Art

[0002] With the rapid development of artificial intelligence (AI) technology, its application in the field of online video processing has become increasingly in-depth, providing users with an unprecedented personalized experience. In online video communication, AI technology is widely used in real-time beautification and transformation of character images, from slight beauty adjustments to deep facial reshaping, and even complex operations such as gender conversion. Although these technologies enhance the fun and personalization of video calls, they also pose a severe challenge to the authenticity of video content. Summary of the invention

[0003] The present application provides a data processing method and an electronic device. The specific scheme is as follows:

[0004] The first aspect of the present application provides a data processing method, comprising:

[0005] The first device obtains target data; the target data is acquired based on the first acquisition module of the first device, and at least includes a first image of the target object, the first image is obtained by processing a second image acquired by the first acquisition module by an image processing module;

[0006] Processing the target data based on a target model to obtain a target processing result; the target model is run on the first device, and the target processing result includes a parameter indicating the identity of the target object;

[0007] The first information is sent to the second device based on the target processing result and the target data, so that the second device outputs the second information based on the first information.

[0008] In a possible implementation, processing the target data based on the target model to obtain a target processing result includes:

[0009] The first image is input into the target model so that the target model outputs a first confidence score; the first confidence score represents the probability that the target object contained in the first image is the object operating the first device, and the greater the first confidence score, the greater the probability that the target object contained in the first image is the object operating the first device.

[0010] In a possible implementation, the target data further includes: a first voice of the target object, the first voice being obtained by processing the second voice collected by the second collection module by the voice processing module;

[0011] The method further comprises:

[0012] The first speech is input into the target model so that the target model outputs a second confidence level; the second confidence level represents the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device, and the greater the second confidence level, the greater the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device.

[0013] In a possible implementation, sending the first information to the second device based on the target processing result and the target data so that the second device outputs the second information based on the first information includes one of the following:

[0014] generating a prompt strategy based on the target processing result, and sending the prompt strategy and the target data to a second device, so that the second device outputs the target data based on the prompt strategy; the first information includes the prompt strategy and the target data, and the second information includes the target data output based on the prompt strategy;

[0015] Sending the target processing result and the target data to a second device, so that the second device generates a prompt strategy based on the target processing result and outputs the target data based on the prompt strategy; the first information includes the target processing result and the target data, and the second information includes the target data output based on the prompt strategy;

[0016] The prompt strategy is used to limit the allowed image processing degree, to issue an alarm when the allowed image processing degree is exceeded, and to output the target data normally when the allowed image processing degree is not exceeded.

[0017] In a possible implementation, generating a prompt strategy based on the target processing result includes any of the following:

[0018] Selecting a target confidence interval to which at least one of the first confidence level and the second confidence level belongs from a plurality of confidence intervals, and using a strategy associated with the target confidence interval as a prompt strategy;

[0019] A target confidence interval to which at least one of the first confidence level and the second confidence level belongs is selected from a plurality of confidence intervals, a type of application to which the target data belongs is determined, and a policy associated with the target confidence interval under the type is used as a prompt policy.

[0020] In a possible implementation, if the type includes: role-playing or game-type video call software, the prompt strategy is used to limit the first image processing degree allowed;

[0021] If the type includes: dating video call software, the prompt strategy is used to limit the allowed second image processing degree;

[0022] If the type includes: live video software, the prompt policy is used to limit the third image processing degree allowed;

[0023] If the type includes: conference video software, the prompt policy is used to limit the fourth image processing degree allowed;

[0024] The second image processing degree is lower than at least one of the first image processing degree, the third image processing degree, and the fourth image processing degree.

[0025] In a possible implementation, the method further includes:

[0026] In the process of the acquisition module continuously acquiring data, a training data set is constructed based on at least part of the data acquired by the acquisition module according to a preset time interval; the data set is used to train the target model; and the data includes at least the second image.

[0027] In a possible implementation, after the first device obtains the target data, the method further includes:

[0028] Authenticating the target data; the authentication includes at least one of the following: determining whether the sender and the generator of the target data are consistent; determining whether the destination receiving address specified by the target data is accurate;

[0029] If the authentication is passed, the step of processing the target data based on the target model to obtain the target processing result is executed.

[0030] In a possible implementation, the processing method further includes:

[0031] Obtaining a target request returned by the second device, where the target request is at least used to request the first device to reduce the difference between the first image and the second image;

[0032] The processing mode of the image processing module on the second image acquired by the first acquisition module is changed based on the target request.

[0033] In another aspect of the present application, an electronic device is provided, comprising:

[0034] at least one of a first acquisition module and a second acquisition module, wherein the first acquisition module is used to acquire a second image, and the second acquisition module is used to acquire a second voice;

[0035] Processor for:

[0036] Obtain target data; the target data is acquired based on the first acquisition module, and includes at least a first image of the target object, the first image being acquired by processing a second image acquired by the first acquisition module by an image processing module;

[0037] Processing the target data based on a target model to obtain a target processing result; the target model is run on the first device, and the target processing result includes a parameter indicating the identity of the target object;

[0038] The first information is sent to the second device based on the target processing result and the target data, so that the second device outputs the second information based on the first information. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0040] Figure 1 A flowchart of a data processing method provided in Example 1 of the present application;

[0041] Figure 2 A flowchart of a data processing method provided in Example 2 of the present application;

[0042] Figure 3 A flowchart of a data processing method provided in Example 3 of the present application;

[0043] Figure 4 A schematic diagram of an implementation scenario of a data processing method provided in this application;

[0044] Figure 5 A schematic diagram of another implementation scenario of a data processing method provided in this application;

[0045] Figure 6 A flowchart of a data processing method provided in Example 7 of the present application;

[0046] Figure 7 A flowchart of a data processing method provided in Example 8 of the present application;

[0047] Figure 8 A schematic diagram of interaction between modules provided in this application;

[0048] Fig. 9 A flowchart of a data processing method provided in Example 9 of the present application;

[0049] Fig.10 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0050] The following describes the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. The terms used in the implementation method section of the present application are only used to explain the specific embodiments of the present application, and are not intended to limit the present application.

[0051] The embodiments of the present application are described below in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0052] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and need not be used to describe a specific order or sequential order. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, which is only to describe the distinction mode adopted by the objects of the same attributes when describing in the embodiments of the present application. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, so that the process, method, system, product or equipment comprising a series of units need not be limited to those units, but may include other units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0053] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0054] Reference Figure 1 , is a flow chart of a data processing method provided in Example 1 of the present application, which can be applied to a first device, such as Figure 1 As shown, the method may include but is not limited to the following steps:

[0055] Step S101, the first device obtains target data; the target data is acquired based on the first acquisition module of the first device, and at least includes a first image of the target object, and the first image is obtained by processing the second image acquired by the first acquisition module by the image processing module.

[0056] In this embodiment, the first acquisition module (eg, a camera or a camera module) of the first device can capture an image of the target object to obtain an original image of the target object, that is, a second image.

[0057] The image processing module can process the second image to obtain the first image of the target object. The image processing module processes the second image, which may include but is not limited to: AI beautification processing such as beauty, face reshaping, and gender conversion.

[0058] The image processing module may be integrated into the system-level function of the first device, rather than being specific to a certain application. The user may manage the enabling and disabling status of the image processing module through the system setting interface of the first device.

[0059] The image processing module can also be deployed in a specific application, and the execution and triggering of the image processing module completely depends on the logic of the application and user operations. In this case, the image processing module will not be directly controlled or managed at the system level, but exists as part of the application function. Users can trigger the execution of the image processing module through the interface or function options within the application. For example, users may need to select options such as "beautify photos" or "apply filters" within the application to instruct the application to call the image processing module to process the image.

[0060] Step S102: Process the target data based on the target model to obtain a target processing result.

[0061] The target model can be run on the first device, and the target model can process the target data locally on the first device, which can ensure the processing efficiency and reduce the risk of target data leakage.

[0062] The target processing result may contain a parameter indicating the identity of the target object.

[0063] The first device may include but is not limited to: a personal computer with integrated AI function. The personal computer with integrated AI function can use powerful computing power and flexible scalability to run the target model to accelerate the processing speed and accuracy of the target model on the target data.

[0064] Of course, the first device may also be other intelligent devices with sufficient computing power and storage resources, such as a smart phone, a tablet computer, etc.

[0065] Step S103: Sending the first information to the second device based on the target processing result and the target data, so that the second device outputs the second information based on the first information.

[0066] The first information may include target data, and data that can directly or indirectly point to the identity of the target object.

[0067] Correspondingly, the second device outputs second information based on the first information, and the second information can enable the user to obtain target data and corresponding prompts, and the prompts can correspond to the identity of the target object.

[0068] In this embodiment, after the image processing module processes the second image acquired by the first acquisition module to obtain the first image of the target object, the first device can obtain the target data, process the target data based on the target model, and obtain the target processing result, which includes parameters indicating the identity of the target object, so as to complete the identification of the identity of the target object in the target data. On this basis, the first information is sent to the second device based on the target processing result and the target data. The first information may include the target data and data that can directly or indirectly point to the identity of the target object. By sending the first information to the second device, the second device can output the second information based on the first information, ensuring that the user can obtain the target data and corresponding prompts by viewing the second information. The prompts can assist the user in understanding the authenticity of the target object, improve the user experience, and enable the user to make judgments or decisions based on more sufficient information.

[0069] As another optional embodiment of the present application, refer to Figure 2 , is a flow chart of a data processing method provided in Example 2 of the present application. In this embodiment, it is mainly an implementation method of step S102 in Example 1, such as Figure 2 As shown, step S102 may include but is not limited to the following steps:

[0070] Step S1021: input the first image into the target model so that the target model outputs a first confidence level; the first confidence level represents the probability that the target object included in the first image is the object for operating the first device.

[0071] In this embodiment, the target model may be trained using images of the object operating the first device, so that the target model can learn the image feature representation of the object operating the first device and make accurate predictions based on these image features.

[0072] After the target model is trained, when the first image is input into the target model, the target model can infer the first image based on the image feature representation it has learned. During the inference process, it can be evaluated whether the target object contained in the first image is the object for operating the first device, and this possibility can be quantified as a first confidence. The first confidence can characterize the probability that the target object contained in the first image is the object for operating the first device.

[0073] Although the image processing module may change some visual attributes of the target object (such as color, brightness, contrast, etc.) when processing the second image, the core content and structure of the target object in the image often remain unchanged or only slightly changed. Therefore, the target model can still use the learned image features to identify the key elements in the first image and obtain the first confidence.

[0074] In this embodiment, the greater the first confidence level, the greater the probability that the target object contained in the first image is the object operating the first device.

[0075] Accordingly, step S103 in Embodiment 1 may include but is not limited to:

[0076] Step S1031: Send first information to a second device based on the first confidence level and target data, so that the second device outputs second information based on the first information.

[0077] The first information may include target data and data that can directly or indirectly point to the first confidence level.

[0078] Correspondingly, the second device outputs second information based on the first information, and the second information allows the user to obtain target data and corresponding prompts, and the prompts may correspond to the first confidence level.

[0079] In this embodiment, a first image is input into a target model so that the target model outputs a first confidence level, and first information is sent to a second device based on the first confidence level and target data. The first information may include target data and data that can directly or indirectly point to the first confidence level. The first information is sent to the second device so that the second device outputs second information based on the first information, thereby ensuring that a user can obtain target data and receive corresponding prompts by viewing the second information. The prompt corresponds to the first confidence level and can intuitively convey to the user the possibility that the target object in the first image is the object for operating the electronic device based on the probability represented by the first confidence level, thereby assisting the user in understanding the authenticity of the target object, improving the user experience, and enabling the user to make judgments or decisions based on more sufficient information.

[0080] As another optional embodiment of the present application, refer to Figure 3 , is a flow chart of a data processing method provided in Example 3 of the present application, such as Figure 3 As shown, the method may include but is not limited to the following steps:

[0081] Step S201, the first device obtains target data; the target data is acquired based on the first acquisition module of the first device, and includes at least a first image and a first voice of the target object, the first image is obtained by processing the second image acquired by the first acquisition module by the image processing module; the first voice is obtained by processing the second voice acquired by the second acquisition module by the voice processing module.

[0082] In this embodiment, the first acquisition module, image processing module, first image and second image in this embodiment can be understood by referring to the relevant introduction of the first acquisition module, image processing module, first image and second image in Embodiment 1, which will not be repeated here.

[0083] The second acquisition module (eg, a microphone or a microphone array) of the first device can acquire the voice of the target object to obtain the original voice of the target object, that is, the second voice.

[0084] The speech processing module can process the second speech to obtain the first speech of the target object. The speech processing module processes the second speech, which may include but is not limited to: changing the sound, speech speed, etc.

[0085] The voice processing module can be integrated into the system-level function of the first device, rather than being specific to a certain application. The user can manage the enabling and disabling status of the voice processing module through the system setting interface of the first device.

[0086] The voice processing module can also be deployed in a specific application, and the execution and triggering of the voice processing module completely depends on the logic and user operation of the application. In this case, the voice processing module is not directly controlled or managed at the system level, but exists as part of the application function. Users can trigger the execution of the voice processing module through the interface or function options within the application. For example, users may need to select options such as "voice change" in the application to instruct the application to call the voice processing module to process the voice.

[0087] Step S202: input the first image into the target model so that the target model outputs a first confidence level; the first confidence level represents the probability that the target object included in the first image is the object for operating the first device.

[0088] The greater the first confidence, the greater the probability that the first image contains the target object that is the object operating the first device.

[0089] In this embodiment, the target model may include but is not limited to: a multimodal model including an image recognition sub-model and a speech recognition sub-model.

[0090] In this embodiment, the first image may be input into the image recognition sub-model so that the image recognition sub-model outputs a first confidence level.

[0091] The detailed process of step S202 can refer to the relevant introduction of step S1021 in embodiment 2, which will not be repeated here.

[0092] Step S203: input the first speech into the target model so that the target model outputs a second confidence level, wherein the second confidence level represents the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device.

[0093] In this step, the first speech may be input into the speech recognition sub-model so that the speech recognition sub-model outputs a second confidence level.

[0094] In this embodiment, the target model may be trained using the voice of the object operating the first device, so that the target model can learn the voiceprint feature representation of the object operating the first device and make accurate predictions based on these voiceprint features.

[0095] After the target model is trained, when the first speech is input into the target model, the target model can infer the first speech based on the voiceprint feature representation it has learned. During the inference process, it can be evaluated whether the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device, and this possibility can be quantified as a second confidence level. The second confidence level can characterize the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device.

[0096] Although the speech processing module may change some speech properties of the target object (such as pitch, timbre, etc.) when processing the second speech, the core content and structure of the target object's speech in the speech often remain unchanged or only change slightly. Therefore, the target model can still use the learned voiceprint features to identify the key elements in the first speech and obtain the second confidence.

[0097] The greater the second confidence level is, the greater the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device.

[0098] Steps S202-S203 are an implementation of step S102 in Example 1.

[0099] Step S204: Send the first information to the second device based on the first confidence level, the second confidence level and the target data, so that the second device outputs the second information based on the first information.

[0100] The first information may include target data, and data that can directly or indirectly point to the first confidence level and the second confidence level.

[0101] Accordingly, the second device outputs second information based on the first information, and the second information can enable the user to obtain target data and corresponding prompts, and the prompts can correspond to the first confidence level and the second confidence level.

[0102] Step S204 is an implementation of step S103 in Example 1.

[0103] In this embodiment, a first image is input into a target model so that the target model outputs a first confidence, a first voice is input into the target model so that the target model outputs a second confidence, and first information is sent to a second device based on the first confidence, the second confidence and the target data. The first information may include target data and data that can directly or indirectly point to the first confidence and the second confidence. The first information is sent to the second device so that the second device outputs the second information based on the first information, thereby ensuring that the user can obtain the target data and receive corresponding prompts by viewing the second information. The prompts correspond to the first confidence and the second confidence, and can intuitively convey to the user the possibility that the target object in the first image is the object for operating the electronic device based on the probabilities represented by the first confidence and the second confidence, thereby providing the user with more comprehensive and in-depth information to evaluate the authenticity of the target object, thereby significantly enhancing the user experience.

[0104] As another optional embodiment of the present application, a data processing method is provided in Embodiment 4 of the present application. This embodiment is mainly an implementation of step S103 in Embodiment 1. Step S103 may include but is not limited to the following:

[0105] Step S1031: Generate a prompt strategy based on the target processing result, and send the prompt strategy and target data to the second device, so that the second device outputs the target data based on the prompt strategy.

[0106] The prompt strategy can be used to limit the allowed image processing level, issue an alarm when the allowed image processing level is exceeded, and output the target data normally when the allowed image processing level is not exceeded.

[0107] The first information may include a prompting strategy and target data, and the second information may include target data output based on the prompting strategy.

[0108] In this embodiment, if the target processing result indicates that the actual image processing degree of the second image acquired by the first acquisition module by the image processing module exceeds the allowable image processing degree, the prompt strategy can be used to limit the allowable image processing degree, and if it exceeds the allowable image processing degree, accordingly, the second device can determine from the prompt strategy that an alarm needs to be issued, and the second device can issue an alarm when outputting target data.

[0109] If the target processing result indicates that the actual image processing degree performed by the image processing module on the second image acquired by the first acquisition module does not exceed the allowable image processing degree, the prompt strategy can be used to limit the allowable image processing degree and does not exceed the allowable image processing degree. Accordingly, the second device can determine from the prompt strategy that it can output the target data normally, and the second device can output the target data normally.

[0110] Step S1032: Send the target processing result and target data to the second device, so that the second device generates a prompt strategy based on the target processing result and outputs the target data based on the prompt strategy; the first information includes the target processing result and target data, and the second information includes the target data output based on the prompt strategy.

[0111] The prompt strategy generated in step S1032 can also be used to limit the allowed image processing degree. When the allowed image processing degree is exceeded, an alarm is issued, and when the allowed image processing degree is not exceeded, the target data is output normally.

[0112] The way in which the second device generates a prompt strategy based on the target processing result can be the same as the method adopted by the first device in step S1031, that is, following the same logic or algorithm to generate the prompt strategy; it can also be different, that is, the second device may adopt a completely new and different logic, algorithm or process to generate a prompt strategy based on the target processing result.

[0113] In this embodiment, after the image processing module processes the second image acquired by the first acquisition module to obtain the first image of the target object, the first device can obtain the target data, process the target data based on the target model, and obtain the target processing result, which includes parameters indicating the identity of the target object, so as to complete the identification of the identity of the target object in the target data. On this basis, the first information is sent to the second device based on the target processing result and the target data. The first information may include the target data and data that can directly or indirectly point to the identity of the target object. By sending the first information to the second device, the second device can output the second information based on the first information, ensuring that the user can obtain the target data and corresponding prompts by viewing the second information. The prompts can assist the user in understanding the authenticity of the target object, improve the user experience, and enable the user to make judgments or decisions based on more sufficient information.

[0114] In this embodiment, a prompt strategy is generated based on the target processing result, and the prompt strategy can directly indicate the operation that the second device should take for the output of the target data, that is, whether to issue an alarm or output normally. If the prompt strategy indicates that the second device should issue an alarm when outputting the target data, the second device can issue an alarm on the image processing degree of the first device while outputting the target data. The alarm can enhance the user's perception of the image processing degree, promote the user's understanding of the authenticity of the target object, and improve the user experience.

[0115] As another optional embodiment of the present application, a data processing method is provided in Embodiment 5 of the present application. This embodiment is mainly an implementation method of generating a prompt strategy based on the target processing result in Embodiment 4. The prompt strategy generated based on the target processing result may include but is not limited to any of the following:

[0116] Step S11: Select a target confidence interval to which at least one of the first confidence level and the second confidence level belongs from a plurality of confidence intervals, and use a strategy associated with the target confidence interval as a prompt strategy.

[0117] In this embodiment, multiple confidence intervals can be set according to actual needs, and the confidence interval can be a continuous or discontinuous numerical range. The numerical range is the range of possible values ​​of the confidence obtained by the target model.

[0118] Multiple confidence intervals may include: a high confidence interval (e.g., greater than 90%), a medium-high confidence interval (e.g., [70%, 90%]), a medium-low confidence interval (e.g., [40%, 70%]), and a low confidence interval (e.g., less than 40%). For example, if the first confidence is in the high confidence interval, it means that the target model believes that the target object in the first image highly matches the features of the object that operates the electronic device. If the first confidence is in the medium-high confidence interval, it means that the target model believes that the target object in the first image has a high degree of match with the features of the object that operates the electronic device, but there may be some uncertainty. If the first confidence is in the medium-low confidence interval, it means that the target model believes that the target object in the first image has a low degree of match with the features of the object that operates the electronic device. If the first confidence is in the low confidence interval, it means that the target model believes that the target object in the first image has a very low degree of match with the features of the object that operates the electronic device.

[0119] Each confidence interval is associated with a policy, and the policies associated with each confidence interval can be different from each other.

[0120] The strategy corresponding to the high confidence interval can be used to indicate that the target object in the first image has hardly been processed by the image processing module or the speech of the target object has hardly been processed by the speech processing module.

[0121] The strategy corresponding to the medium and high confidence intervals can be used to indicate that the target object in the first image has been processed to a certain extent by the image processing module or that the speech of the target object has been processed to a certain extent by the speech processing module.

[0122] The strategy corresponding to the medium and low confidence intervals can be used to indicate that the target object in the first image has been processed to a large extent by the image processing module or that the speech of the target object has been processed to a large extent by the speech processing module.

[0123] The strategy corresponding to the low confidence interval can be used to indicate that the target object in the first image has been over-processed by the image processing module or that the speech of the target object has been over-processed by the speech processing module.

[0124] For example, the second image Figure 4As shown in part (a), if the target confidence interval to which the first confidence level belongs is a medium-high confidence interval, the second device outputs the first image based on the prompt strategy. Figure 4 As shown in part (b), Figure 4 As shown in part (b), the second device may output a prompt message, namely, "the image has a certain degree of beautification", while outputting the first image.

[0125] For another example, the second image is Figure 4 As shown in part (a), if the target confidence intervals to which the first confidence level and the second confidence level belong are both low confidence intervals, the second device outputs the first image based on the prompt strategy. Figure 5 ,like Figure 5 As shown, the second device can output prompt information while outputting the first image and the first voice, namely, "the image is over-beautified" and "the voice does not match the age and gender".

[0126] In this embodiment, based on the refined policy management method of associating each confidence interval with its own policy, a target confidence interval to which at least one of the first confidence level and the second confidence level belongs is selected from multiple confidence intervals, and the policy associated with the target confidence interval is used as the prompt policy. The accuracy of the prompt policy can be guaranteed, and the second device can output the target data based on the prompt policy, so that the user can obtain more accurate prompts, understand the authenticity of the target object more accurately, and enhance the user experience.

[0127] Step S12: Select a target confidence interval to which at least one of the first confidence level and the second confidence level belongs from multiple confidence intervals, determine the type of application to which the target data belongs, and use a policy associated with the target confidence interval under the type as a prompt policy.

[0128] The setting of multiple confidence intervals may refer to the related introduction of setting multiple confidence intervals in step S11, which will not be repeated here.

[0129] In step S12, each confidence interval may be associated with at least one policy, and at least one policy corresponds to at least one type of application program. The same confidence interval may be associated with different policies under different types of application programs.

[0130] For example, the low confidence interval can correspond to four policies, and the types of applications corresponding to the four policies are a, b, c and d respectively. If the target confidence interval is a low confidence interval, and the type of application to which the target data belongs is a, the policy associated with the low confidence interval under type a is used as the prompt policy. The prompt policy can be used to limit the image processing degree allowed under type a, and issue an alarm when the image processing degree allowed under type a is exceeded, and output the target data normally when the image processing degree allowed under type a is not exceeded.

[0131] In this embodiment, after determining the application type to which the target data belongs, a strategy associated with the type (i.e., a strategy associated with the target confidence interval under the type) can be provided as a prompt strategy in a targeted manner. According to the customized prompt strategy, the target data can be output more reasonably and accurately, thereby reducing the output error caused by the use of a general prompt strategy.

[0132] As another optional embodiment of the present application, a data processing method is provided for Example 6 of the present application. This embodiment is mainly an implementation method of the type and prompt strategy in Example 5. The type of application to which the target data belongs may include but is not limited to: role-playing or game video call software; or, dating video call software; or, video live broadcast software; or, conference video software.

[0133] If the type includes: role-playing or game-type video call software, the prompt strategy is used to limit the allowed first image processing degree, to issue an alarm when the allowed first image processing degree is exceeded, and to output the target data normally when the allowed first image processing degree is not exceeded;

[0134] If the type includes: dating video call software, the prompt strategy is used to limit the allowed second image processing degree, and to issue an alarm when the allowed second image processing degree is exceeded, and to output the target data normally when the allowed second image processing degree is not exceeded;

[0135] If the type includes: video live broadcast software, the prompt strategy is used to limit the allowed third image processing degree, to issue an alarm when the allowed third image processing degree is exceeded, and to output the target data normally when the allowed third image processing degree is not exceeded;

[0136] If the type includes: conference video software, the prompt strategy is used to limit the fourth image processing degree allowed, and to issue an alarm when the fourth image processing degree is exceeded, and to output the target data normally when the fourth image processing degree is not exceeded;

[0137] In this embodiment, for role-playing or game-type video call software, the first image processing degree can be set relatively high, that is, a certain degree of image processing (such as beautification and deformation) is allowed to meet the special needs of users in games or role-playing. When the degree of processing of the target object in the second image by the image processing module is within the first image processing degree, the prompt strategy can be used to prompt the normal output of the target data.

[0138] For dating video call software, the second image processing degree can be lower than at least one of the first image processing degree, the third image processing degree and the fourth image processing degree. Because dating video call software has relatively high requirements for the authenticity of the target object in the first image, the second image processing degree can be set relatively low, that is, when the image processing module performs slight processing on the second image (such as the target object and other areas other than the target object in the first image), the prompt strategy should also be able to prompt that the second image processing degree allowed is exceeded and an alarm is required. For example, the prompt strategy can be used to prompt that the target object has been beautified to a corresponding degree, and that the background part other than the target object has been modified.

[0139] For live video software, a certain degree of beautification is allowed but excessive beautification is not encouraged. Therefore, the third image processing degree can be set relatively high. When the image processing module processes the second image to a degree exceeding the third image processing degree, the prompt strategy can prompt that the third image processing degree is exceeded. For example, the prompt strategy can be used to prompt that the target object is enhanced by AI beautification technology, and that the background part other than the target object is modified.

[0140] For conference video software, the main focus is on communication during the meeting, and the requirements for whether the first image or the first voice belongs to the object of operating the electronic device are not very high. Therefore, the fourth image processing level can be set relatively high. When the image processing module processes the second image to a degree that exceeds the fourth image processing level, the prompt strategy can prompt cautious communication to ensure the effectiveness of the meeting.

[0141] As another optional embodiment of the present application, refer to Figure 6 , is a flow chart of a data processing method provided in Example 7 of the present application, such as Figure 6 As shown, the method may include but is not limited to the following steps:

[0142] Step S301: During the process of continuous data collection by the collection module, a training data set is constructed based on at least part of the data collected by the collection module according to a preset time interval; the data set is used to train the target model.

[0143] In this embodiment, the acquisition module may include at least a first acquisition module. Of course, the acquisition module may also include at least one of the second acquisition modules. The data collected by the first acquisition module and the second acquisition module can be referred to the introduction in the above-mentioned related embodiments, and will not be repeated here.

[0144] Accordingly, the data may at least include the second image acquired by the first acquisition module.

[0145] The preset time interval can be set as needed and is not limited in this application.

[0146] The target model is trained based on the original image, that is, the second image, and can learn more basic and essential features in the second image. These features still maintain a certain stability after the image processing module transforms the image. Therefore, when the target model processes the first image, the accuracy and robustness of the processing can be guaranteed.

[0147] Step S302, the first device obtains target data; the target data is acquired based on the first acquisition module of the first device, and at least includes a first image of the target object, and the first image is obtained by processing the second image acquired by the first acquisition module by the image processing module.

[0148] Step S303: Process the target data based on the target model to obtain a target processing result; the target model runs on the first device, and the target processing result includes parameters indicating the identity of the target object.

[0149] Step S304: Send first information to the second device based on the target processing result and the target data, so that the second device outputs second information based on the first information.

[0150] The detailed process of steps S302-S304 can refer to the relevant introduction of steps S101-S103 in Example 1, which will not be repeated here.

[0151] In this embodiment, during the process of continuous data collection by the acquisition module, a training data set is constructed based on at least part of the data collected by the acquisition module according to a preset time interval; the data set is used to train the target model, and the target model infers the authenticity of the target object in the first image through the learned feature representation, rather than relying on a direct comparison between the first image and the original image (such as the second image). This processing method is more difficult to be predicted and bypassed by counterfeiters. In addition, there is no need to store the original image (such as the second image) locally on the first device, which can avoid leakage of the original image.

[0152] As another optional embodiment of the present application, refer to Figure 7 , is a flow chart of a data processing method provided in Example 8 of the present application, such as Figure 7 As shown, the method may include but is not limited to the following steps:

[0153] Step S401, the first device obtains target data; the target data is acquired based on the first acquisition module of the first device, and at least includes a first image of the target object, and the first image is obtained by processing the second image acquired by the first acquisition module by the image processing module.

[0154] The detailed process of step S401 can refer to the relevant introduction of step S101 in embodiment 1, which will not be repeated here.

[0155] Step S402, authenticating the target data; the authentication includes at least one of the following: determining whether the sender and the generator of the target data are consistent; determining whether the destination receiving address specified by the target data is accurate.

[0156] In this embodiment, the authentication logic can be encapsulated into an independent data transmission authentication interface. The authentication logic can include at least one of the following: determining whether the sender and the generator of the target data are consistent; determining whether the destination receiving address specified by the target data is accurate.

[0157] After obtaining the target data, the data sending authentication interface can be called, and according to the result returned by the data sending authentication interface, it can be decided whether to continue processing the target data.

[0158] Step S403: If the authentication is successful, the target data is processed based on the target model to obtain a target processing result; the target model is run on the first device, and the target processing result includes parameters indicating the identity of the target object.

[0159] If the result returned by the data sending authentication interface is that the authentication is passed, the target data can be processed based on the target model to obtain the target processing result.

[0160] Step S404: Send first information to the second device based on the target processing result and the target data, so that the second device outputs second information based on the first information.

[0161] In this embodiment, by determining whether the sender of the target data is consistent with the generator, it is possible to ensure that the target data is not used or tampered with by other video software. By verifying the accuracy of the destination receiving address, it is possible to prevent the target data from being sent to the wrong recipient, thereby preventing the data from being leaked to unauthorized third parties.

[0162] In this embodiment, the authentication logic can be encapsulated into an independent data transmission authentication interface to make the authentication process more standardized and easier to manage. Figure 8 As shown, the original data (such as the second image and the second voice) can be captured by the original data capture module, the processing module can process the original data to obtain the target data (such as the first image and the first voice), and before sending the target data to the second device, the data processing module can call the data sending authentication interface to authenticate the target data. If the result returned by the data sending authentication interface is that the authentication is passed, the data processing module processes the target data based on the target model to obtain the target processing result, and sends the first information to the second device based on the target processing result and the target data through the network module.

[0163] Among them, the target model can run on the AIPC NPU (Neural Processing Unit) to accelerate processing and improve processing accuracy.

[0164] The raw data can also be used as input to a raw data analysis module. The raw data analysis module constructs a training data set based on at least part of the raw data at a preset time interval, and trains a target model based on the data set.

[0165] As another optional embodiment of the present application, refer to Fig. 9 , is a flow chart of a data processing method provided in Example 9 of the present application, such as Fig. 9 As shown, the method may include but is not limited to the following steps:

[0166] Step S501, the first device obtains target data; the target data is acquired based on the first acquisition module of the first device, and at least includes a first image of the target object, and the first image is obtained by processing the second image acquired by the first acquisition module by the image processing module.

[0167] Step S502: Process the target data based on the target model to obtain a target processing result; the target model runs on the first device, and the target processing result includes parameters indicating the identity of the target object.

[0168] Step S503: Send the first information to the second device based on the target processing result and the target data, so that the second device outputs the second information based on the first information.

[0169] The detailed process of steps S501-S503 can refer to the relevant introduction of steps S101-S102 in Example 1, which will not be repeated here.

[0170] Step S504: Obtain a target request returned by the second device, where the target request is at least used to request the first device to reduce the difference between the first image and the second image.

[0171] The second device may return a target request when running a specific type of application program, for example, when running a dating video call software.

[0172] The second device may return the target request through the data sending authentication interface.

[0173] By returning the target request, the first device can perform customized processing.

[0174] Step S505: changing the processing mode of the image processing module for the second image acquired by the first acquisition module based on the target request.

[0175] Changing the processing mode may include, but is not limited to, changing at least one of parameter settings of an image processing module and an image processing algorithm.

[0176] After the processing mode is changed, the difference between the first image obtained by processing the second image and the second image can be reduced.

[0177] In this embodiment, by obtaining the target request returned by the second device and changing the processing mode of the image processing module on the second image captured by the first acquisition module based on the target request, the output first information can meet user expectations and further improve the user experience.

[0178] Next, a data processing device provided by the present application is introduced. The data processing device introduced below and the data processing method introduced above can refer to each other.

[0179] The data processing means may include:

[0180] A first acquisition unit is used to obtain target data; the target data is acquired based on the first acquisition module of the first device, and at least includes a first image of the target object, the first image is obtained by processing the second image acquired by the first acquisition module by the image processing module;

[0181] A processing unit, configured to process the target data based on the target model to obtain a target processing result; the target model is run on the first device, and the target processing result includes a parameter indicating the identity of the target object;

[0182] The sending unit is used to send the first information to the second device based on the target processing result and the target data, so that the second device outputs the second information based on the first information.

[0183] The processing unit can be used for:

[0184] The first image is input into the target model so that the target model outputs a first confidence level; the first confidence level represents the probability that the target object contained in the first image is the object operating the first device, and the greater the first confidence level, the greater the probability that the target object contained in the first image is the object operating the first device.

[0185] The target data may also include: a first voice of the target object, the first voice being obtained by processing the second voice collected by the second collection module by the voice processing module;

[0186] Processing units can also be used to:

[0187] The first speech is input into the target model so that the target model outputs a second confidence level; the second confidence level represents the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device, and the greater the second confidence level, the greater the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device.

[0188] The sending unit can be used to:

[0189] Generate a prompt strategy based on the target processing result, and send the prompt strategy and target data to the second device, so that the second device outputs the target data based on the prompt strategy; the first information includes the prompt strategy and the target data, and the second information includes the target data output based on the prompt strategy;

[0190] or,

[0191] Sending a target processing result and target data to a second device, so that the second device generates a prompt strategy based on the target processing result and outputs the target data based on the prompt strategy; the first information includes the target processing result and the target data, and the second information includes the target data output based on the prompt strategy;

[0192] The prompt strategy is used to limit the allowed image processing level. When the allowed image processing level is exceeded, an alarm is issued. When the allowed image processing level is not exceeded, the target data is output normally.

[0193] The process of the sending unit generating a prompt strategy based on the target processing result may specifically include any of the following:

[0194] Selecting a target confidence interval to which at least one of the first confidence level and the second confidence level belongs from a plurality of confidence intervals, and using a strategy associated with the target confidence interval as a prompt strategy;

[0195] A target confidence interval to which at least one of the first confidence level and the second confidence level belongs is selected from a plurality of confidence intervals, a type of application to which the target data belongs is determined, and a policy associated with the target confidence interval under the type is used as a prompt policy.

[0196] If the type includes: role-playing or gaming video call software, the prompt policy is used to limit the first image processing degree allowed;

[0197] If the type includes: dating video call software, the prompt policy is used to limit the degree of second image processing allowed;

[0198] If the type includes: video live streaming software, the prompt policy is used to limit the degree of third-party image processing allowed;

[0199] If the type includes: video conferencing software, the prompt policy is used to limit the fourth image processing degree allowed;

[0200] The second image processing degree is lower than at least one of the first image processing degree, the third image processing degree, and the fourth image processing degree.

[0201] The data processing device may further include:

[0202] A construction unit is used to construct a training data set based on at least part of the data collected by the acquisition module according to a preset time interval during the process of the acquisition module continuously collecting data; the data set is used to train the target model; the data includes at least a second image.

[0203] The data processing device may further include:

[0204] An authentication unit is used to authenticate the target data; the authentication includes at least one of the following: determining whether the sender and the generator of the target data are consistent; determining whether the destination receiving address specified by the target data is accurate;

[0205] If the authentication is passed, the step of processing the target data based on the target model to obtain the target processing result is executed.

[0206] The data processing device may further include:

[0207] A second obtaining unit, configured to obtain a target request returned by the second device, wherein the target request is at least used to request the first device to reduce the difference between the first image and the second image;

[0208] The adjustment unit is used to change the processing mode of the image processing module on the second image acquired by the first acquisition module based on the target request.

[0209] In another embodiment of the present application, an electronic device is provided, referring to Fig.10 , the electronic device includes: a collection module 100 and a processor 200.

[0210] The acquisition module 100 includes: at least one of a first acquisition module and a second acquisition module, the first acquisition module is used to acquire a second image, and the second acquisition module is used to acquire a second voice;

[0211] The processor 200 is configured to:

[0212] Obtain target data; the target data is acquired based on the first acquisition module, and at least includes a first image of the target object, the first image being acquired by processing a second image acquired by the first acquisition module by an image processing module;

[0213] Processing the target data based on the target model to obtain a target processing result; the target model is run on the first device, and the target processing result includes a parameter indicating the identity of the target object;

[0214] The first information is sent to the second device based on the target processing result and the target data, so that the second device outputs the second information based on the first information.

[0215] It should also be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the device embodiments provided by the present application, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines.

[0216] Through the description of the above implementation mode, the technicians in the field can clearly understand that the present application can be implemented by means of software plus necessary general hardware, and of course, it can also be implemented by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. In general, all functions completed by computer programs can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better implementation mode in more cases. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a readable storage medium, such as a computer floppy disk, a U disk, a mobile hard disk, a ROM, a RAM, a disk or an optical disk, etc., including a number of instructions to enable a computer device (which can be a personal computer, a training device, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0217] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0218] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a training device, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, training device, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device, a data center, etc. that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

Claims

1. A data processing method, include: The first device obtains target data; The target data is acquired based on the first acquisition module of the first device, and includes at least a first image of the target object, the first image being acquired by processing a second image acquired by the first acquisition module by an image processing module; Processing the target data based on the target model to obtain a target processing result; The target model is run on the first device, and the target processing result includes a parameter indicating the identity of the target object; The first information is sent to the second device based on the target processing result and the target data, so that the second device outputs the second information based on the first information.

2. The data processing method according to claim 1, processing the target data based on the target model to obtain a target processing result, include: Inputting the first image into the target model so that the target model outputs a first confidence level; The first confidence value represents the probability that the target object contained in the first image is the object operating the first device. The greater the first confidence value, the greater the probability that the target object contained in the first image is the object operating the first device.

3. The data processing method according to claim 2, wherein the target data is include: The first speech of the target object is obtained by processing the second speech collected by the second collection module by the speech processing module; The method further comprises: Inputting the first speech into the target model so that the target model outputs a second confidence level; The second confidence level represents the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device. The greater the second confidence level, the greater the probability that the voiceprint feature of the target object contained in the first speech belongs to the object operating the first device.

4. The data processing method according to any one of claims 1 to 3, wherein the step of sending first information to a second device based on the target processing result and the target data so that the second device outputs second information based on the first information comprises one of the following: generating a prompt strategy based on the target processing result, and sending the prompt strategy and the target data to a second device, so that the second device outputs the target data based on the prompt strategy; the first information includes the prompt strategy and the target data, and the second information includes the target data output based on the prompt strategy; Sending the target processing result and the target data to a second device, so that the second device generates a prompt strategy based on the target processing result and outputs the target data based on the prompt strategy; the first information includes the target processing result and the target data, and the second information includes the target data output based on the prompt strategy; The prompt strategy is used to limit the allowed image processing degree, to issue an alarm when the allowed image processing degree is exceeded, and to output the target data normally when the allowed image processing degree is not exceeded.

5. The data processing method according to claim 4, wherein the prompt strategy is generated based on the target processing result, including any of the following: Selecting a target confidence interval to which at least one of the first confidence level and the second confidence level belongs from a plurality of confidence intervals, and using a strategy associated with the target confidence interval as a prompt strategy; A target confidence interval to which at least one of the first confidence level and the second confidence level belongs is selected from a plurality of confidence intervals, a type of application to which the target data belongs is determined, and a policy associated with the target confidence interval under the type is used as a prompt policy.

6. The data processing method according to claim 5, if the type include: Role-playing or gaming video call software, wherein the prompt strategy is used to limit the first image processing degree allowed; If the type includes: dating video call software, the prompt strategy is used to limit the allowed second image processing degree; If the type includes: live video software, the prompt policy is used to limit the third image processing degree allowed; If the type includes: conference video software, the prompt policy is used to limit the fourth image processing degree allowed; The second image processing degree is lower than at least one of the first image processing degree, the third image processing degree, and the fourth image processing degree.

7. The data processing method according to claim 1, further comprising: include: In the process of the acquisition module continuously acquiring data, constructing a training data set based on at least part of the data acquired by the acquisition module according to a preset time interval; The data set is used to train the target model; the data at least includes the second image.

8. The data processing method according to claim 1, after the first device obtains the target data, include: authenticating the target data; The authentication includes at least one of the following: determining whether the sender and the generator of the target data are consistent; Determining whether the destination receiving address specified by the target data is accurate; If the authentication is passed, the step of processing the target data based on the target model to obtain the target processing result is executed.

9. The data processing method according to claim 1, further comprising: include: Obtaining a target request returned by the second device, where the target request is at least used to request the first device to reduce the difference between the first image and the second image; The processing mode of the image processing module on the second image acquired by the first acquisition module is changed based on the target request.

10. An electronic device, include: A collection module, the collection module comprising: at least one of a first collection module and a second collection module, the first collection module is used to collect a second image, and the second collection module is used to collect a second voice; Processor for: Obtain target data; the target data is acquired based on the first acquisition module, and includes at least a first image of the target object, the first image being acquired by processing a second image acquired by the first acquisition module by an image processing module; Processing the target data based on a target model to obtain a target processing result; the target model is run on the first device, and the target processing result includes a parameter indicating the identity of the target object; The first information is sent to the second device based on the target processing result and the target data, so that the second device outputs the second information based on the first information.