Live exposure adjustment method and device, equipment and medium thereof

Abstract

The application discloses a live exposure adjustment method and device, equipment and medium, the method comprises the following steps: in response to a first exposure setting instruction, performing exposure adjustment and face recognition on a first live picture, obtaining reference face information and a reference exposure value; in response to a second exposure setting instruction, performing face recognition on a second live picture to obtain current face information, and performing face difference detection on the current face information based on the reference face information; when the detection result is that the current face information has face difference, performing exposure value adjustment on the second live picture based on the reference exposure value, taking the adjusted exposure value as the reference exposure value, and identifying the latest face information as the reference face information. The application provides a live picture exposure adjustment based on faces for a network live platform, so as to optimize the clarity of the live picture and improve the live viewing experience of users.

Description

Technical Field

[0001] This application relates to the field of live streaming, and more particularly to a method for adjusting live streaming exposure. It also relates to the corresponding apparatus, equipment, and non-volatile storage medium. Background Technology

[0002] Existing online live streaming platforms feature various types of live streamers. Each streamer broadcasts their recorded live stream to their respective live streaming room, allowing viewers to receive and play the stream on their devices. The live stream is then displayed on the live streaming interface, enabling viewers to understand the streamer's content. However, the exposure value of the live stream recorded by the streamer determines the brightness of the final displayed live stream. Excessive or insufficient exposure will affect the clarity of the live stream. For example, excessive exposure will result in overly bright live streams, making it difficult for viewers to clearly see the content. Conversely, insufficient exposure will cause the same problem. Therefore, controlling the exposure value of the live stream is a crucial issue that online live streaming platforms need to address.

[0003] In view of the problems with the exposure value of existing live broadcast images, the applicant has made corresponding explorations in order to solve the problem. Summary of the Invention

[0004] The purpose of this application is to provide a live broadcast exposure adjustment method to meet user needs, and also relates to the corresponding apparatus, equipment, non-volatile storage medium and computer program products.

[0005] To achieve the objectives of this application, the following technical solution is adopted:

[0006] A live streaming exposure adjustment method proposed for the purposes of this application includes the following steps:

[0007] In response to the first exposure setting command, the system adjusts the exposure of the first live broadcast image and performs facial recognition to obtain baseline facial information and baseline exposure value.

[0008] In response to the second exposure setting command, face recognition is performed on the second live broadcast to obtain the current face information, and face difference detection is performed on the current face information based on the reference face information;

[0009] When the detection result indicates that there is a facial difference in the current facial information, the exposure value of the second live broadcast image is adjusted based on the baseline exposure value, the adjusted exposure value is used as the baseline exposure value, and the latest facial information is identified as the baseline facial information.

[0010] In a further embodiment, the step of responding to the first exposure setting command, performing face recognition on the first live broadcast image, and obtaining reference face information and reference exposure value includes the following steps:

[0011] Acquire live screen image frames from the first live screen, and generate multiple live screen image frames with different exposure values ​​according to the exposure value adjustment rules;

[0012] The face recognition model trained to convergence is invoked to identify the face information in each of the live broadcast image frames;

[0013] The target live stream frame with the most complete facial information is identified among all the live stream image frames. The facial information and exposure value of the target live stream frame are used as the reference facial information and reference exposure value, respectively.

[0014] In a further embodiment, the step of responding to the first exposure setting command, performing face recognition on the first live broadcast image, and obtaining reference face information and reference exposure value includes the following steps:

[0015] In response to an exposure value determination instruction applied to the screen exposure value adjustment control, the exposure value applied to the first live screen corresponding to the exposure value determination instruction is obtained;

[0016] The face recognition model trained to convergence is invoked to identify the face information in the first live broadcast frame corresponding to the exposure value;

[0017] The facial information and the exposure value are used as the reference facial information and the reference exposure value, respectively.

[0018] In a further embodiment, the step of performing facial difference detection on the current facial information based on the reference facial information includes the following steps:

[0019] Obtain multiple baseline facial feature data contained in the baseline facial information;

[0020] Determine whether the current face information contains the baseline face feature data. If not, output a detection result indicating that there is a face difference.

[0021] In a further embodiment, the step of adjusting the exposure value of the second live broadcast image based on the reference exposure value, using the adjusted exposure value as the reference exposure value, and identifying the latest facial information as the reference facial information includes the following steps:

[0022] Based on the baseline exposure value, it is determined whether the latest exposure value adjusted for the second live broadcast image is overexposed;

[0023] When the latest exposure value exceeds the baseline exposure value, a notification indicating that the exposure value is too high will be displayed on the current live broadcast interface.

[0024] When the latest exposure value is less than the baseline exposure value, a notification indicating that the exposure value is too low will be displayed on the current live broadcast interface.

[0025] The system obtains the determined exposure value from the screen exposure value adjustment control as the baseline exposure value, and identifies the facial information in the live screen after the exposure value is adjusted as the baseline facial information.

[0026] In a further embodiment, the step of adjusting the exposure value of the second live broadcast image based on the reference exposure value, using the adjusted exposure value as the reference exposure value, and identifying the latest facial information as the reference facial information includes the following steps:

[0027] Acquire live screen image frames from the second live screen, determine the exposure value adjustment range based on the reference exposure value, and generate multiple live screen image frames with different exposure values ​​according to the exposure value adjustment range;

[0028] The face recognition model trained to convergence is invoked to identify the face information in each of the live broadcast image frames;

[0029] The target live stream frame with the most complete facial information is identified among all the live stream image frames. The facial information and exposure value of the target live stream frame are used as the reference facial information and reference exposure value, respectively.

[0030] In a further embodiment, after the step of identifying the latest facial information as the reference facial information, the following steps are included:

[0031] The face recognition model, trained to convergence, is invoked to perform face recognition on the latest live stream footage in real time.

[0032] When a face cannot be identified from the latest live stream footage, a notification indicating that no face exists in the current live stream will be displayed on the current live stream interface.

[0033] A live broadcast exposure adjustment device proposed for the purposes of this application includes:

[0034] The baseline information acquisition module is used to respond to the first exposure setting command, adjust the exposure of the first live broadcast screen and perform face recognition to acquire baseline face information and baseline exposure value;

[0035] The face difference detection module is used to respond to the second exposure setting command, perform face recognition on the second live broadcast to obtain the current face information, and perform face difference detection on the current face information based on the reference face information;

[0036] The exposure value adjustment module is used to adjust the exposure value of the second live broadcast image based on the baseline exposure value when the detection result indicates that there is a facial difference in the current facial information. The adjusted exposure value is used as the baseline exposure value, and the latest facial information is identified as the baseline facial information.

[0037] In a further embodiment, the benchmark information acquisition module includes:

[0038] The image frame exposure generation submodule is used to acquire live screen image frames in the first live screen and generate multiple live screen image frames with different exposure values ​​according to the exposure value adjustment rules.

[0039] The face information recognition submodule is used to call the face recognition model trained to convergence and identify the face information in each of the live broadcast image frames;

[0040] The reference information acquisition submodule is used to determine the target live broadcast frame with the most complete facial information in each of the live broadcast image frames, and to use the facial information and exposure value of the target live broadcast frame as the reference facial information and reference exposure value, respectively.

[0041] In a preferred embodiment, the reference information acquisition module further includes:

[0042] The exposure value acquisition submodule is used to respond to the exposure value determination instruction applied to the screen exposure value adjustment control and acquire the exposure value applied to the first live screen corresponding to the exposure value determination instruction.

[0043] The face information recognition submodule is used to call the face recognition model trained to convergence and identify the face information in the first live broadcast frame corresponding to the exposure value;

[0044] The reference information determination submodule is used to use the face information and the exposure value as reference face information and reference exposure value, respectively.

[0045] In a further embodiment, the face difference detection module includes:

[0046] The facial feature data acquisition submodule is used to acquire multiple benchmark facial feature data contained in the benchmark facial information;

[0047] The face information detection submodule is used to determine whether the current face information contains the reference face feature data. If it does not contain them, the module outputs a detection result indicating that there is a face difference.

[0048] In a further embodiment, the exposure value adjustment module includes:

[0049] The overexposure judgment submodule is used to determine, based on the baseline exposure value, whether the latest exposure value adjusted for the second live broadcast image is overexposed;

[0050] The overexposure alert submodule is used to display an alert notification indicating that the exposure value is too high in the current live broadcast interface when the latest exposure value exceeds the baseline exposure value.

[0051] The low exposure alert submodule is used to display a low exposure alert notification on the current live stream interface when the latest exposure value is less than the baseline exposure value.

[0052] The baseline information determination submodule is used to obtain the determined exposure value from the screen exposure value adjustment control as the baseline exposure value, and to identify the facial information in the live screen after adjustment based on the exposure value as the baseline facial information.

[0053] In a preferred embodiment, the exposure value adjustment module further includes:

[0054] The image frame exposure generation submodule is used to acquire live screen image frames in the second live screen, determine the exposure value adjustment range based on the reference exposure value, and generate multiple live screen image frames with different exposure values ​​according to the exposure value adjustment range.

[0055] The face information recognition submodule is used to call the face recognition model trained to convergence and identify the face information in each of the live broadcast image frames;

[0056] The reference information acquisition submodule is used to determine the target live broadcast frame with the most complete facial information in each of the live broadcast image frames, and to use the facial information and exposure value of the target live broadcast frame as the reference facial information and reference exposure value, respectively.

[0057] To address the aforementioned technical problems, this application also provides a computer device, including a memory and a processor. The memory stores computer-readable instructions, which, when executed by the processor, cause the processor to perform the steps of the aforementioned live broadcast exposure adjustment method.

[0058] To address the aforementioned technical problems, this application also provides a storage medium storing computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of the live broadcast exposure adjustment method described above.

[0059] To address the aforementioned technical problems, this application also provides a computer program product, including a computer program and computer instructions. When the computer program and computer instructions are executed by a processor, the processor performs the steps of the aforementioned live broadcast exposure adjustment method.

[0060] Compared with existing technologies, the advantages of this application are as follows:

[0061] This application provides a function for adjusting the exposure value of live stream images on online platforms. The exposure value is adjusted based on the streamer's face displayed in the live stream to improve the clarity of the live content. When a streamer first activates the live stream exposure value adjustment function, the exposure value of the current live stream image is adjusted to ensure the streamer's face is clearly displayed. The adjusted exposure value and the identified complete facial features are used as the baseline exposure value and baseline facial information, respectively. In subsequent exposure value adjustments, the feature comparison between the latest identified facial information and the baseline facial information determines whether the streamer's face in the latest live stream image is clear. If there are differences in facial features between the latest identified facial information and the baseline facial information, the exposure value is adjusted based on the baseline exposure value. The system adjusts the exposure value of the latest live stream to ensure that the streamer's face is clearly displayed. When providing the streamer with an exposure adjustment control, it compares a baseline exposure value with the streamer's current adjustment, displaying an overexposure warning to indicate whether the adjusted exposure is sufficient to fully display the streamer's face. This application uses facial information from the live stream as a reference for exposure adjustment, ensuring the adjusted live stream fully displays the streamer's face, optimizing the clarity and recognizability of the streamer's face, and improving the viewing experience. A better viewing experience can effectively increase viewers' watch time and retention rate in the live stream. Attached Figure Description

[0062] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0063] Figure 1 A typical network deployment architecture diagram related to the implementation of the technical solution of this application;

[0064] Figure 2 This is a flowchart illustrating a typical embodiment of the live broadcast exposure adjustment method of this application;

[0065] Figure 3 This is a graphical user interface diagram of the live streaming room interface that displays the screen exposure value adjustment control in this application;

[0066] Figure 4 This is a graphical user interface diagram of the live streaming room interface that displays a notification indicating that the exposure value is too high, as described in this application.

[0067] Figure 5 A graphical user interface diagram of the live streaming room interface that displays a notification indicating that the exposure value is too low, as described in this application.

[0068] Figure 6 This is a flowchart illustrating the specific implementation method of automatically adjusting the exposure value of the first live broadcast and recognizing facial information in the first live broadcast in this application.

[0069] Figure 7 This is a flowchart illustrating the specific implementation method of adjusting the exposure value of the first live broadcast screen using an exposure value adjustment control in this application.

[0070] Figure 8 This is a flowchart illustrating the specific implementation method for comparing the latest identified current face information with the baseline face information in this application;

[0071] Figure 9 This is a flowchart illustrating a specific implementation of the present application regarding the display of different overexposure warning notifications based on a reference exposure value.

[0072] Figure 10 This is a flowchart illustrating the specific implementation method of automatically adjusting the exposure value of the second live broadcast image based on a benchmark exposure value and identifying facial information in the second live broadcast image in this application.

[0073] Figure 11 This is a graphical user interface diagram of the live streaming room interface that displays a notification that no face is present in this application.

[0074] Figure 12 This is a flowchart illustrating the specific implementation method for providing a notification that no human face is present in the live stream footage, as described in this application.

[0075] Figure 13 This is a schematic block diagram of a typical embodiment of the live broadcast exposure adjustment device of this application;

[0076] Figure 14 This is a basic structural block diagram of a computer device according to an embodiment of this application. Detailed Implementation

[0077] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0078] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this application means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or wireless coupling. The term “and / or” as used herein includes all or any units and all combinations of one or more associated listed items.

[0079] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0080] Those skilled in the art will understand that the terms "client," "terminal," and "terminal device" as used herein include both devices that receive wireless signals, devices that only possess wireless signal receiver capabilities without transmission capabilities, and devices with receiving and transmitting hardware, devices that have receiving and transmitting hardware capable of bidirectional communication over a bidirectional communication link. Such devices may include: cellular or other communication devices such as personal computers or tablets, having single-line displays, multi-line displays, or cellular or other communication devices without multi-line displays; PCS (Personal Communications Service) that can combine voice, data processing, fax, and / or data communication capabilities; PDA (Personal Digital Assistant) that may include a radio frequency receiver, pager, internet / intranet access, web browser, notepad, calendar, and / or GPS (Global Positioning System) receiver; and traditional laptops and / or handheld computers or other devices that have and / or include radio frequency receivers. As used herein, "client," "terminal," and "terminal device" can be portable, transportable, installed in a means of transportation (air, sea, and / or land), or suitable and / or configured to operate locally and / or in a distributed manner, operating in any other location on Earth and / or in space. "Client," "terminal," and "terminal device" as used herein can also be a communication terminal, an internet access terminal, or a music / video playback terminal, such as a PDA, a MID (Mobile Internet Device), and / or a mobile phone with music / video playback capabilities, or a smart TV, set-top box, etc.

[0081] The hardware referred to by the names "server," "client," and "work node" in this application is essentially an electronic device with the equivalent capabilities of a personal computer. It is a hardware device with the necessary components revealed by the von Neumann architecture, such as a central processing unit (including an arithmetic logic unit and a control unit), memory, input devices, and output devices. The computer program is stored in its memory, and the central processing unit loads the program stored in the secondary storage into the main memory to run it, executes the instructions in the program, and interacts with the input and output devices to complete specific functions.

[0082] It should be noted that the concept of "server" used in this application can also be extended to the case of server clusters. Based on the network deployment principles understood by those skilled in the art, the servers should be logically divided. Physically, these servers can be independent of each other but accessible through interfaces, or they can be integrated into a single physical computer or a computer cluster. Those skilled in the art should understand this flexibility and should not use it to constrain the implementation of the network deployment method in this application.

[0083] Please see Figure 1 The hardware infrastructure required for implementing the technical solutions of this application can be deployed according to the architecture shown in the figure. The server 80 mentioned in this application is deployed in the cloud and acts as an online server. It can further connect to relevant data servers and other servers providing related support, thereby forming a logically related service cluster to provide services to relevant terminal devices such as the smartphone 81 and personal computer 82 shown in the figure, or third-party servers (not shown). Both the smartphone and personal computer can access the Internet through known network access methods and establish a data communication link with the cloud server 80 to run terminal applications related to the services provided by the server.

[0084] For servers, the application is usually built as a service process, with corresponding program interfaces exposed for remote calls by applications running on various terminal devices. The relevant technical solutions in this application that are suitable for running on servers can be implemented in servers in this way.

[0085] The application mentioned refers to an application running on a server or terminal device. This application implements the relevant technical solutions of this application in a programmed manner. Its program code can be stored in a non-volatile storage medium that can be recognized by a computer in the form of computer-executable instructions, and is loaded into memory by the central processing unit for execution. The relevant device of this application is constructed by the operation of the application on the computer.

[0086] For servers, the application is usually built as a service process, with corresponding program interfaces exposed for remote calls by applications running on various terminal devices. The relevant technical solutions in this application that are suitable for running on servers can be implemented in servers in this way.

[0087] Those skilled in the art will understand that although the various methods in this application are described based on the same concept and thus present commonality among them, they can be performed independently unless otherwise specified. Similarly, the various embodiments disclosed in this application are all based on the same inventive concept; therefore, concepts expressed in the same way, as well as concepts that are appropriately changed for convenience but are expressed differently, should be understood equivalently.

[0088] Please see Figure 2 A live streaming exposure adjustment method according to this application, in a typical embodiment, includes the following steps:

[0089] Step S11: In response to the first exposure setting command, adjust the exposure and perform face recognition on the first live broadcast image to obtain baseline face information and baseline exposure value.

[0090] The first exposure setting command is generally triggered by the broadcaster's client to adjust the exposure value of the live stream recorded by the broadcaster's client. For example, if the first live stream is the live stream currently displayed in the broadcaster's client's live stream interface, the broadcaster triggers the first exposure setting command through the corresponding controls in the live stream interface to adjust the exposure value of the first live stream.

[0091] The exposure value refers to the numerical value used to modulate the exposure of the live broadcast image. Generally, the exposure value of the live broadcast image recorded by the device's camera can be set. Live broadcast images with different exposure values ​​will have different image brightness when displayed on the live broadcast interface. By adjusting the exposure value of the live broadcast image, the content displayed in the live broadcast image can be made clearer, and the clarity and recognizability of the face of the broadcaster in the live broadcast image can be improved, so that the viewers can clearly see the face of the broadcaster in the live broadcast image.

[0092] When performing face recognition on the first live stream, a face recognition model that has been trained to convergence will be invoked. The face recognition model is a model built based on a neural network, such as using MTCNN to build the face recognition model, in order to identify the facial shape feature data, left eye feature data, right eye feature data, nose feature data, or mouth feature data of the face displayed in the first live stream. These facial feature data will be encapsulated into the face information of the first live stream.

[0093] The method for adjusting the exposure value of the first live broadcast image can be an automatic exposure value adjustment method. Specifically, the host user terminal currently displaying the first live broadcast image obtains the live broadcast image frames in the first live broadcast image, and generates multiple live broadcast image frames with different exposure values ​​according to the exposure value adjustment rules. Then, it calls the face recognition model trained to convergence to identify the face information in each of the live broadcast image frames, and determines the target live broadcast image frame with the most complete face information in each of the live broadcast image frames. Then, the face information and exposure value of the target live broadcast image frame are used as the reference face information and reference exposure value, respectively.

[0094] Please refer to Figure 3 The method for adjusting the exposure value of the first live stream image can be that the broadcaster can adjust the exposure value using the image exposure value adjustment control. Specifically, Figure 3 The screen exposure value adjustment control 302 shown in the live broadcast interface is used to adjust the screen exposure value. When the broadcaster determines the corresponding exposure value through the screen exposure value adjustment control 302, the broadcaster user terminal will respond to the exposure value determination instruction acting on the screen exposure value adjustment control, obtain the exposure value acting on the first live broadcast screen corresponding to the exposure value determination instruction, and call the face recognition model trained to convergence to identify the face information in the first live broadcast screen corresponding to the exposure value. Then, the face information and the exposure value are used as the reference face information and the reference exposure value, respectively.

[0095] It is understandable that when the exposure value in the first live stream is automatically adjusted by the broadcaster's client, the baseline exposure value is the exposure value corresponding to the live stream frame with the most facial feature data, and the baseline facial information is the facial information corresponding to the face displayed in the live stream frame; when the exposure value in the first live stream is adjusted by the broadcaster on the broadcaster's client, the baseline exposure value is the exposure value determined by the broadcaster through the screen exposure value adjustment control, and the baseline facial information is the facial information present in the first live stream using the baseline exposure value; it can be seen that, in addition to the broadcaster's client automatically adjusting the exposure value to reduce the broadcaster's operational needs, the broadcaster can also adjust the exposure value independently through the control to meet the broadcaster's customized needs for the display effect of the live stream.

[0096] Step S12: In response to the second exposure setting command, perform face recognition on the second live stream to obtain current face information, and perform face difference detection on the current face information based on the baseline face information.

[0097] The second exposure setting command generally refers to the exposure setting command triggered after responding to the first exposure setting command. For example, in some scenarios, when the anchor user starts live streaming debugging or during the live streaming process, the second exposure setting command will generally be triggered to readjust the exposure of the currently displayed second live streaming screen so that the live streaming content can be clearly displayed in the second live streaming screen, such as improving the clarity of the anchor user's face in the second live streaming screen.

[0098] The second live stream screen refers to the live stream screen displayed in the current live stream interface when responding to the second exposure setting command.

[0099] After the broadcaster responds to the second exposure setting command, the broadcaster's client will perform facial difference detection on the facial information in the second live broadcast based on the baseline facial information. Specifically, it will call the face recognition model that has been trained to convergence, identify the current facial information corresponding to the currently displayed face in the second live broadcast, and obtain multiple facial feature data contained in the baseline facial information to determine whether the current facial information contains all the facial feature data contained in the baseline facial information. If the current facial information contains all the facial feature data, it indicates that the current exposure value of the second live broadcast can clearly display the broadcaster's face, and the corresponding output will show a detection result that there is no facial difference in the current facial information. If the current facial information does not contain all the facial feature data, it indicates that the current exposure value of the second live broadcast cannot clearly display the broadcaster's face, and the corresponding output will show a detection result that there is a facial difference in the current facial information.

[0100] When it is detected that there is no facial difference between the current facial information in the second live broadcast and the reference facial information, the exposure value corresponding to the second exposure setting instruction will be obtained. For example, the exposure value determined by the broadcast user on the broadcast user end through the screen exposure value adjustment control displayed in the live broadcast interface is the exposure value corresponding to the second exposure setting instruction. This exposure value is used as the latest reference exposure value, and the current facial information obtained from the second live broadcast by the face recognition is used as the latest reference exposure value. The screen exposure value adjustment control can refer to the screen exposure value adjustment control 302 shown in Figure 3.

[0101] Step S13: When the detection result indicates that there is a facial difference in the current facial information, the exposure value of the second live broadcast image is adjusted based on the baseline exposure value. The adjusted exposure value is used as the baseline exposure value, and the latest facial information is identified as the baseline facial information.

[0102] When a difference is detected between the current facial information in the second live stream and the baseline facial information, the exposure value of the second live stream will be adjusted based on the baseline exposure value so that the face of the broadcaster in the second live stream can be fully displayed.

[0103] Please refer to Figure 4 and Figure 5 When the exposure value of the second live stream is adjusted by the broadcaster through the exposure value adjustment control, such as when the broadcaster adjusts the exposure value through... Figure 4 The image exposure adjustment control 401 shown is... Figure 5 When the exposure value adjustment control 501 shown adjusts the exposure value of the second live broadcast screen, the broadcaster's client will provide a corresponding overexposure warning based on the baseline exposure value. Specifically, the broadcaster's client will determine whether the latest exposure value of the second live broadcast screen is overexposed based on the baseline exposure value. When the latest exposure value exceeds the baseline exposure value, a notification indicating that the exposure value is too high will be displayed on the current live broadcast interface. The notification is as follows: Figure 4 The notification 401, indicating an excessively high exposure value, displays a low exposure value notification on the current live stream interface when the latest exposure value is lower than the baseline exposure value. The notification is as follows: Figure 5 The notification 501, which indicates that the exposure value is too high, means that after the broadcaster user determines the exposure value to be adjusted through the screen exposure value adjustment control, the exposure value will be used as the latest baseline exposure value, and the facial information in the live broadcast screen after the exposure value adjustment will be identified as the latest baseline facial information.

[0104] In addition, when the exposure value of the second live broadcast is adjusted automatically based on the reference exposure value, the broadcaster's client will acquire live broadcast image frames in the second live broadcast, use the reference exposure value as the median to determine the exposure value adjustment range, generate multiple live broadcast image frames with different exposure values ​​within the exposure value adjustment range, and then call the face recognition model trained to convergence to identify the face information in each of the live broadcast image frames, determine the target live broadcast frame with the most complete face information in each of the live broadcast image frames, adjust the exposure value of the second live broadcast to the exposure value of the target live broadcast frame, and use this exposure value as the latest reference exposure value, and identify the face information in the live broadcast after the exposure value adjustment is completed as the latest reference face information.

[0105] As can be seen from the typical implementation of this method, it can provide network platforms with live broadcast exposure value adjustment. Based on the broadcaster's face displayed in the live broadcast, the exposure value is adjusted to improve the clarity of the live content. When the broadcaster first activates the live broadcast exposure value adjustment function, the exposure value of the current live broadcast is adjusted to clearly display the broadcaster's face. The currently adjusted exposure value and the identified complete facial features are used as the baseline exposure value and baseline facial information, respectively. In subsequent exposure value adjustments, the feature comparison between the latest identified facial information and the baseline facial information determines whether the broadcaster's face in the latest live broadcast is clear. If there are differences in facial features between the latest identified facial information and the baseline facial information... The system adjusts the exposure value of the latest live stream based on a baseline exposure value to ensure that the streamer's face is clearly displayed. When providing the streamer with an exposure adjustment control, the system compares the baseline exposure value with the streamer's current adjustment, displaying an overexposure warning to indicate whether the adjusted exposure value is sufficient to fully display the streamer's face. Essentially, this method uses the facial information in the live stream as a reference for exposure adjustment, ensuring that the adjusted live stream fully displays the streamer's face, optimizing the clarity and recognizability of the streamer's face, and improving the viewing experience in the live stream room. A better viewing experience can effectively increase the viewers' watch time and retention rate in the live stream room.

[0106] The above typical embodiments and their variations fully disclose the implementation scheme of the live broadcast exposure adjustment method of this application. However, various variations of the method can still be derived by changing and expanding some technical means. Other embodiments are briefly described below:

[0107] In one embodiment, please refer to Figure 6 The step of responding to the first exposure setting command, performing face recognition on the first live broadcast screen, and obtaining reference face information and reference exposure value includes the following steps:

[0108] Step S111: Obtain live video image frames from the first live video feed, and generate multiple live video image frames with different exposure values ​​according to the exposure value adjustment rules.

[0109] The exposure value rules set multiple different exposure values. The range of each exposure value is generally determined based on the basic exposure value preset by the developers. The preset basic exposure value is used as the median to set the range of values, and each exposure value in the exposure value rules is determined. Then, whenever the live broadcast image frame is adjusted to the corresponding exposure value in the exposure value rules, a live broadcast image frame with the currently adjusted exposure value will be generated, so as to generate multiple live broadcast image frames with different exposure values.

[0110] Step S112: Invoke the face recognition model trained to convergence to identify the face information in each of the live broadcast image frames:

[0111] The face recognition model is a model built based on a neural network. For example, the face recognition model is built based on MTCNN to identify the face shape feature data, left eye feature data, right eye feature data, nose feature data or mouth feature data displayed in each of the live broadcast image frames as the face information, and generate the face information corresponding to each of the live broadcast image frames.

[0112] Step S113: Determine the target live stream frame with the most complete facial information among all the live stream image frames, and use the facial information and exposure value of the target live stream frame as the reference facial information and reference exposure value, respectively.

[0113] By comparing the facial feature data contained in each of the identified facial information, the facial information with the most facial feature data is selected as the most complete facial information, and the live broadcast image frame corresponding to the facial information is used as the target live broadcast frame. Then, the facial information and exposure value of the target live broadcast frame are used as the reference facial information and reference exposure value, respectively.

[0114] When there are two or more facial information sets, the live video frame with the highest or lowest exposure value among the corresponding live video frames is selected as the target live video frame.

[0115] In this embodiment, after responding to the first exposure setting command, the exposure value of the first live broadcast screen will be automatically adjusted. Based on the completeness of the facial information in the first live broadcast screen under each exposure value, the exposure value corresponding to the most complete facial information is determined as the benchmark exposure value, and the most complete facial information is used as the benchmark facial information. This eliminates the need for the broadcaster to manually adjust the screen exposure value, thereby improving the live broadcast user experience.

[0116] In one embodiment, please refer to Figure 3 and 7The step of responding to the first exposure setting command, performing face recognition on the first live broadcast screen, and obtaining reference face information and reference exposure value includes the following steps:

[0117] Step S111': In response to the exposure value determination instruction applied to the screen exposure value adjustment control, obtain the exposure value applied to the first live screen corresponding to the exposure value determination instruction.

[0118] Please refer to Figure 3 , Figure 3 The screen exposure value adjustment control 302 shown in the live broadcast interface can be used to adjust the screen exposure value. When the broadcaster determines the corresponding exposure value through the screen exposure value adjustment control 302, the broadcaster's client will respond to the exposure value determination instruction acting on the screen exposure value adjustment control and obtain the exposure value determined by the broadcaster in the screen exposure value adjustment control 302.

[0119] Step S112': Invoke the face recognition model trained to convergence to identify the face information in the first live broadcast frame corresponding to the exposure value:

[0120] The facial features, left eye features, right eye features, nose features, or mouth features displayed in the live image frame of the first live broadcast that has been adjusted to the exposure value are identified as the facial information.

[0121] Step S113': The facial information and the exposure value are respectively used as the reference facial information and the reference exposure value.

[0122] After identifying the facial information in the first live stream frame that has been adjusted to the exposure value determined by the broadcaster user through the screen exposure value adjustment control, the facial information is used as the reference facial information, and the determined exposure value is set as the reference exposure value.

[0123] In this embodiment, a screen exposure value adjustment control is provided to the broadcaster, so that the broadcaster can customize and adjust the screen exposure value to match the current live broadcast content, thereby improving the viewing experience of the live broadcast.

[0124] In one embodiment, please refer to Figure 8 The step of performing facial difference detection on the current facial information based on the reference facial information includes the following steps:

[0125] Step S121: Obtain multiple reference face feature data contained in the reference face information:

[0126] The data types of the facial feature data can be divided into facial shape feature data, left eye feature data, right eye feature data, nose feature data, and mouth feature data. The baseline facial feature data contained in the baseline facial information generally includes facial feature data of all data types. That is, the baseline facial information generally includes facial shape feature data, left eye feature data, right eye feature data, nose feature data, and mouth feature data that can represent the complete face of the anchor user.

[0127] Step S122: Determine whether the current face information contains each of the baseline face feature data. If not, output a detection result indicating that a face difference exists.

[0128] Based on the aforementioned baseline facial feature data, it is determined whether the latest identified current facial information in the second live broadcast image contains the aforementioned baseline facial feature data. If it does, it indicates that the current exposure value of the second live broadcast image can clearly display the face of the broadcaster user, and a detection result indicating that there is no facial difference in the current facial information is output. If it does not, it indicates that the current exposure value of the second live broadcast image cannot clearly display the face of the broadcaster user, and a detection result indicating that there is a facial difference in the current facial information is output.

[0129] In this embodiment, based on the baseline face information with complete facial feature data, it is determined whether the latest identified face information in the second live broadcast frame currently undergoing exposure adjustment has complete face adjustment data of the baseline face information, so as to determine whether the face displayed in the second live broadcast frame currently undergoing exposure value adjustment is clear, and then the exposure value of the second live broadcast frame is adjusted. The face information of the live broadcast frame is used as the reference object for exposure value adjustment, so that the live broadcast frame after exposure value adjustment can completely display the face of the anchor user.

[0130] In one embodiment, please refer to Figures 3 to 5 and Figure 9 The step of adjusting the exposure value of the second live broadcast image based on the reference exposure value, using the adjusted exposure value as the reference exposure value, and identifying the latest facial information as the reference facial information includes the following steps:

[0131] Step S131: Based on the baseline exposure value, determine whether the latest exposure value adjusted for the second live stream is overexposed.

[0132] Please refer to Figure 3 The latest exposure value is generally obtained by the streamer user on the streamer's client side through... Figure 3The exposure value determined by the exposure value adjustment control 302 shown in the image is compared with the reference exposure value to determine whether the latest exposure value causes the second live broadcast image to be overexposed.

[0133] Step S132: When the latest exposure value exceeds the baseline exposure value, a notification indicating that the exposure value is too high is displayed on the current live stream interface.

[0134] Please refer to Figure 4 When the latest exposure value exceeds the baseline exposure value, a notification indicating that the exposure value is too high will be displayed on the current live stream interface. The notification is as follows: Figure 4 The displayed warning notification 401 indicates that the exposure value is too high.

[0135] Step S133: When the latest exposure value is less than the baseline exposure value, a notification indicating that the exposure value is too low is displayed on the current live stream interface.

[0136] Please refer to Figure 5 When the latest exposure value is less than the baseline exposure value, a notification indicating that the exposure value is too low will be displayed on the current live stream interface. The notification is as follows: Figure 5 The displayed warning message 501 indicates that the exposure value is too high.

[0137] Step S134: Obtain the determined exposure value from the image exposure value adjustment control as the reference exposure value, and identify the facial information in the live broadcast image after adjustment based on the exposure value as the reference facial information:

[0138] Please refer to Figure 3 , Figure 3 The screen exposure value adjustment control 302 shown in the live broadcast interface can be used to adjust the screen exposure value. When the broadcaster determines the corresponding exposure value through the screen exposure value adjustment control 302, the broadcaster's client will respond to the exposure value determination instruction acting on the screen exposure value adjustment control, obtain the exposure value determined by the broadcaster in the screen exposure value adjustment control 302 as the latest baseline exposure value, and at the same time identify the facial information in the live broadcast screen adjusted based on the exposure value determined in the screen exposure value adjustment control 302 as the latest baseline facial information.

[0139] In this embodiment, when the broadcaster adjusts the exposure value of the live broadcast screen using the control, the baseline exposure value is compared with the exposure value adjusted by the broadcaster using the control. Different overexposure prompts are displayed in the live broadcast interface to inform the broadcaster of the current exposure value adjustment status, so as to provide the broadcaster with a reference for adjusting the exposure value and improve the efficiency of exposure value adjustment.

[0140] In one embodiment, please refer to Figure 10 The step of adjusting the exposure value of the second live broadcast image based on the reference exposure value, using the adjusted exposure value as the reference exposure value, and identifying the latest facial information as the reference facial information includes the following steps:

[0141] Step S131': Acquire live screen image frames from the second live screen, determine the exposure value adjustment range based on the reference exposure value, and generate multiple live screen image frames with different exposure values ​​according to the exposure value adjustment range.

[0142] The special effects frames of the live broadcast displayed in the second live broadcast are obtained, and the exposure value adjustment range is determined based on the reference exposure value. The exposure value adjustment range is determined by using the reference exposure value as the median, and multiple exposure values ​​are determined within the exposure value adjustment range. Then, the live broadcast image frames are adjusted according to these exposure values ​​to generate multiple live broadcast image frames with different exposure values.

[0143] Step S132': Invoke the face recognition model trained to convergence to identify the face information in each of the live broadcast image frames:

[0144] The face recognition model is a model built based on a neural network. For example, the face recognition model is built based on MTCNN to identify the face shape feature data, left eye feature data, right eye feature data, nose feature data or mouth feature data displayed in each of the live broadcast image frames as the face information, and generate the face information corresponding to each of the live broadcast image frames.

[0145] Step S133': Determine the target live stream frame with the most complete facial information among all the live stream image frames, and use the facial information and exposure value of the target live stream frame as the reference facial information and reference exposure value, respectively.

[0146] By comparing the facial feature data contained in each of the identified facial information, the facial information with the most facial feature data is selected as the most complete facial information, and the live broadcast image frame corresponding to the facial information is used as the target live broadcast frame. Then, the facial information and exposure value of the target live broadcast frame are used as the reference facial information and reference exposure value, respectively.

[0147] When there are two or more facial information sets, the live video frame with the highest or lowest exposure value among the corresponding live video frames is selected as the target live video frame.

[0148] In this embodiment, when a difference is detected between the facial information in the second live broadcast and the baseline facial information, the exposure value of the first live broadcast will be automatically adjusted. Based on the completeness of the facial information in the first live broadcast under each exposure value, the exposure value corresponding to the most complete facial information is determined as the baseline exposure value. The most complete facial information is used as the baseline facial information, eliminating the need for the broadcaster to manually adjust the exposure value and improving the live broadcast user experience.

[0149] In one embodiment, please refer to Figure 11 and Figure 12 After the step of identifying the latest facial information as the reference facial information, the following steps are included:

[0150] Step S14: Call the face recognition model that has been trained to convergence and perform face recognition on the latest live stream footage in real time:

[0151] When the broadcaster's client enables the exposure value adjustment function, the broadcaster's client will call the face recognition model in real time to perform real-time or timed face recognition processing on the live broadcast screen captured by the broadcaster's client, so as to obtain the latest face information displayed in the live broadcast screen, and then adjust the exposure value of the live broadcast screen based on the face information.

[0152] Step S15: When facial information cannot be identified from the latest live stream footage, a notification indicating that no face exists in the current live stream interface will be displayed.

[0153] Please refer to Figure 11 , Figure 11 If no face is displayed in the live stream, the face recognition model will be unable to identify the face information from the displayed live stream. In this case, the live stream interface will display something like... Figure 11 The indication shown is notification 1101, indicating that there is no human face in the image.

[0154] In this embodiment, when adjusting the exposure value of the live broadcast screen, if the face of the broadcaster is not displayed in the live broadcast screen, a prompt notification will be displayed in the live broadcast interface indicating that there is no face in the screen, so as to prompt the broadcaster to place the face in the live broadcast screen.

[0155] Furthermore, by functionalizing the various steps in the methods disclosed in the above embodiments, a live broadcast exposure adjustment device of this application can be constructed. Following this approach, please refer to... Figure 13In one typical embodiment, the device includes: a reference information acquisition module 11, configured to respond to a first exposure setting command, adjust the exposure of a first live broadcast image and perform face recognition to acquire reference face information and a reference exposure value; a face difference detection module 12, configured to respond to a second exposure setting command, perform face recognition on a second live broadcast image to acquire current face information, and perform face difference detection on the current face information based on the reference face information; and an exposure value adjustment module 13, configured to adjust the exposure value of the second live broadcast image based on the reference exposure value when the detection result indicates that there is a face difference in the current face information, use the adjusted exposure value as the reference exposure value, and identify the latest face information as the reference face information.

[0156] In one embodiment, the reference information acquisition module 11 includes: an image frame exposure generation submodule, used to acquire live screen image frames in the first live screen and generate multiple live screen image frames with different exposure values ​​according to exposure value adjustment rules; a face information recognition submodule, used to call a face recognition model trained to convergence and identify the face information in each of the live screen image frames; and a reference information acquisition submodule, used to determine the target live screen frame with the most complete face information in each of the live screen image frames, and use the face information and exposure value of the target live screen frame as reference face information and reference exposure value, respectively.

[0157] In another embodiment, the reference information acquisition module 11 further includes: an exposure value acquisition submodule, configured to respond to an exposure value determination instruction acting on the screen exposure value adjustment control, and acquire the exposure value acting on the first live screen corresponding to the exposure value determination instruction; a face information recognition submodule, configured to call a face recognition model trained to convergence, and identify the face information present in the first live screen corresponding to the exposure value; and a reference information determination submodule, configured to use the face information and the exposure value as reference face information and reference exposure value, respectively.

[0158] In one embodiment, the face difference detection module 12 includes: a face feature data acquisition submodule, used to acquire multiple reference face feature data contained in the reference face information; and a face information detection submodule, used to determine whether the current face information contains each of the reference face feature data, and if not, output a detection result indicating that a face difference exists.

[0159] In one embodiment, the exposure value adjustment module 13 includes: an overexposure judgment submodule, used to determine whether the latest exposure value adjusted for the second live broadcast image is overexposed based on the baseline exposure value; an overexposure prompt submodule, used to display a prompt notification indicating overexposure in the current live broadcast interface when the latest exposure value exceeds the baseline exposure value; an underexposure prompt submodule, used to display a prompt notification indicating underexposure in the current live broadcast interface when the latest exposure value is less than the baseline exposure value; and a baseline information determination submodule, used to obtain a determined exposure value from the image exposure value adjustment control as the baseline exposure value, and identify facial information in the live broadcast image adjusted based on the exposure value as the baseline facial information.

[0160] In another embodiment, the exposure value adjustment module 13 further includes: an image frame exposure generation submodule, used to acquire live screen image frames in the second live screen, determine an exposure value adjustment range based on the reference exposure value, and generate multiple live screen image frames with different exposure values ​​according to the exposure value adjustment range; a face information recognition submodule, used to call a face recognition model trained to convergence, and identify the face information in each of the live screen image frames; and a reference information acquisition submodule, used to determine the target live screen frame with the most complete face information in each of the live screen image frames, and use the face information and exposure value of the target live screen frame as reference face information and reference exposure value, respectively.

[0161] To address the aforementioned technical problems, this application also provides a computer device for running a computer program implemented according to the live broadcast exposure adjustment method. Please refer to the following for details. Figure 14 , Figure 14 This is a basic structural block diagram of the computer device in this embodiment.

[0162] like Figure 14 The diagram shows the internal structure of a computer device. This computer device includes a processor, non-volatile storage medium, memory, and a network interface connected via a system bus. The non-volatile storage medium stores an operating system, a database, and computer-readable instructions. The database may store a sequence of control information. When executed by the processor, the computer-readable instructions enable the processor to implement a live exposure adjustment method. The processor provides computational and control capabilities, supporting the operation of the entire computer device. The memory stores computer-readable instructions, which, when executed by the processor, enable the processor to implement a live exposure adjustment method. The network interface of the computer device is used for communication with a terminal. Those skilled in the art will understand that… Figure 14The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0163] In this embodiment, the processor executes the specific functions of each module / submodule in the live exposure adjustment device of this application, and the memory stores the program code and various types of data required to execute the above modules. The network interface is used for data transmission between the user terminal and the server. In this embodiment, the memory stores the program code and data required to execute all modules / submodules in the live exposure adjustment device, and the server can call the server's program code and data to execute the functions of all submodules.

[0164] This application also provides a non-volatile storage medium in which the live exposure adjustment method is written as a computer program and stored in the storage medium in the form of computer-readable instructions. When the computer-readable instructions are executed by one or more processors, it means that the program is running in the computer, thereby causing one or more processors to perform the steps of the live exposure adjustment method of any of the above embodiments.

[0165] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0166] In summary, this application provides face-based exposure adjustment for live streaming platforms to optimize the clarity of the live stream and improve the user's viewing experience.

[0167] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0168] Those skilled in the art will understand that the steps, measures, and solutions in the various operations, methods, and processes discussed in this application can be alternated, modified, combined, or deleted. Furthermore, other steps, measures, and solutions in the various operations, methods, and processes discussed in this application can also be alternated, modified, rearranged, decomposed, combined, or deleted. Furthermore, steps, measures, and solutions in the prior art that are similar to those disclosed in this application can also be alternated, modified, rearranged, decomposed, combined, or deleted.

[0169] The above description is only a partial embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for adjusting live stream exposure, characterized in that, Includes the following steps: In response to the first exposure setting command, the system adjusts the exposure of the first live broadcast image and performs facial recognition to obtain baseline facial information and baseline exposure value. In response to the second exposure setting command, face recognition is performed on the second live broadcast to obtain the current face information, and face difference detection is performed on the current face information based on the reference face information; When the detection result indicates that there is a facial difference in the current facial information, the exposure value adjustment range is determined based on the baseline exposure value. Multiple candidate live broadcast frames with different exposure values ​​are generated according to the exposure value adjustment range. The facial recognition model is called to identify the facial information in each candidate live broadcast frame. The target live broadcast frame with the most complete facial information is selected. The exposure value of the target live broadcast frame is used as the new baseline exposure value, and the facial information of the target live broadcast frame is used as the new baseline facial information.

2. The method according to claim 1, characterized in that, The step of responding to the first exposure setting command, performing face recognition on the first live broadcast screen, and obtaining the baseline face information and baseline exposure value includes the following steps: Acquire live screen image frames from the first live screen, and generate multiple live screen image frames with different exposure values ​​according to the exposure value adjustment rules; The face recognition model trained to convergence is invoked to identify the face information in each of the live broadcast image frames; The target live stream frame with the most complete facial information is identified among all the live stream image frames. The facial information and exposure value of the target live stream frame are used as the reference facial information and reference exposure value, respectively.

3. The method according to claim 1, characterized in that, The step of responding to the first exposure setting command, performing face recognition on the first live broadcast screen, and obtaining the baseline face information and baseline exposure value includes the following steps: In response to an exposure value determination instruction applied to the screen exposure value adjustment control, the exposure value applied to the first live screen corresponding to the exposure value determination instruction is obtained; The face recognition model trained to convergence is invoked to identify the face information in the first live broadcast frame corresponding to the exposure value; The facial information and the exposure value are used as the reference facial information and the reference exposure value, respectively.

4. The method according to claim 1, characterized in that, The step of performing facial difference detection on the current facial information based on the reference facial information includes the following steps: Obtain multiple baseline facial feature data contained in the baseline facial information; Determine whether the current face information contains the baseline face feature data. If not, output a detection result indicating that there is a face difference.

5. The method according to claim 1, characterized in that, The steps of determining the exposure value adjustment range based on the baseline exposure value, generating multiple candidate live broadcast frames with different exposure values ​​according to the exposure value adjustment range, calling a face recognition model to identify the face information in each candidate live broadcast frame, selecting the target live broadcast frame with the most complete face information, using the exposure value of the target live broadcast frame as the new baseline exposure value, and using the face information of the target live broadcast frame as the new baseline face information, include the following steps: Based on the baseline exposure value, it is determined whether the latest exposure value adjusted for the second live broadcast image is overexposed; When the latest exposure value exceeds the baseline exposure value, a notification indicating that the exposure value is too high will be displayed on the current live broadcast interface. When the latest exposure value is less than the baseline exposure value, a notification indicating that the exposure value is too low will be displayed on the current live broadcast interface. The determined exposure value is obtained from the exposure value adjustment control as the baseline exposure value, and the facial information in the live broadcast frame after adjustment based on the exposure value is identified as the baseline facial information.

6. The method according to claim 1, characterized in that, The steps of determining the exposure value adjustment range based on the baseline exposure value, generating multiple candidate live broadcast frames with different exposure values ​​according to the exposure value adjustment range, calling a face recognition model to identify the face information in each candidate live broadcast frame, selecting the target live broadcast frame with the most complete face information, using the exposure value of the target live broadcast frame as the new baseline exposure value, and using the face information of the target live broadcast frame as the new baseline face information, include the following steps: Acquire live screen image frames from the second live screen, determine the exposure value adjustment range based on the reference exposure value, and generate multiple live screen image frames with different exposure values ​​according to the exposure value adjustment range; The face recognition model trained to convergence is invoked to identify the face information in each of the live broadcast image frames; The target live stream frame with the most complete facial information is identified among all the live stream image frames. The facial information and exposure value of the target live stream frame are used as the reference facial information and reference exposure value, respectively.

7. The method according to claim 1, characterized in that, After the step of calling the face recognition model to identify the face information in each candidate live broadcast frame, the following steps are included: The face recognition model, trained to convergence, is invoked to perform face recognition on the latest live frame in real time. When a face cannot be identified from the latest live stream frame, a notification indicating that no face exists in the current live stream will be displayed on the current live stream interface.

8. A live broadcast exposure adjustment device, characterized in that, include: The baseline information acquisition module is used to respond to the first exposure setting command, adjust the exposure of the first live broadcast screen and perform face recognition to acquire baseline face information and baseline exposure value; The face difference detection module is used to respond to the second exposure setting command, perform face recognition on the second live broadcast to obtain the current face information, and perform face difference detection on the current face information based on the reference face information; The exposure value adjustment module is used to determine the exposure value adjustment range based on the baseline exposure value when the detection result indicates that there is a facial difference in the current facial information. Based on the exposure value adjustment range, multiple candidate live broadcast frames with different exposure values ​​are generated. The facial recognition model is called to identify the facial information in each candidate live broadcast frame. The target live broadcast frame with the most complete facial information is selected, and the exposure value of the target live broadcast frame is used as the new baseline exposure value. The facial information of the target live broadcast frame is also used as the new baseline facial information.

9. An electronic device comprising a central processing unit and a memory, characterized in that, The central processing unit is used to invoke and run a computer program stored in the memory to perform the steps of the method as described in any one of claims 1 to 7.

10. A non-volatile storage medium, characterized in that, It stores, in the form of computer-readable instructions, a computer program implemented according to any one of claims 1 to 7, which, when invoked by a computer, performs the steps included in the method.

Images