Shooting guidance information processing method and device, electronic equipment and computer readable storage medium

Abstract

The embodiment of the invention discloses a shooting guidance information processing method and device, electronic equipment and a computer readable storage medium. The method comprises the steps of obtaining shooting description information; searching a first theme which is semantically matched with the shooting description information; obtaining visual information corresponding to the first theme; and outputting shooting guidance information corresponding to the first theme according to the visual information. Therefore, according to the scheme, the shooting guidance information corresponding to the theme matched with the shooting description information can be provided for the user, and the user is guided to shoot the theme.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of photographing technology, and specifically to a method, device, electronic device and computer-readable storage medium for processing photographing guidance information. Background Art

[0002] Smart cameras have functions such as auto focus, auto exposure, white balance and smart composition. However, since most users lack professional photography skills and aesthetic knowledge, the quality of the photos they take is often uneven. Taking photos with themes like professional photographers requires more creativity and ideas from users. Therefore, for most users, it is more difficult to take photos of the themes they want. Summary of the invention

[0003] The embodiments of the present application provide a method, device, electronic device and computer-readable storage medium for processing shooting guidance information, which can provide users with shooting guidance information corresponding to a theme matching the shooting description information, and guide users to perform theme shooting.

[0004] In a first aspect, an embodiment of the present application provides a method for processing shooting guidance information, including:

[0005] Get the shooting description information;

[0006] Finding a first subject that semantically matches the shooting description information;

[0007] Acquiring visual information corresponding to the first theme; and

[0008] Based on the visual information, shooting guidance information corresponding to the first theme is output.

[0009] In one embodiment, obtaining visual information corresponding to the first theme includes:

[0010] The first topic is input into a topic learning model to obtain the visual information.

[0011] In one embodiment, the training step of the topic learning model at least includes:

[0012] Obtaining image samples of each subject sample, wherein the image samples carry visual annotation information;

[0013] Inputting the picture samples of each theme sample into the initial theme learning model to obtain the visual prediction information corresponding to each theme sample output by the initial theme learning model;

[0014] The initial topic learning model is trained according to the difference between the visual prediction information and the visual annotation information corresponding to each topic sample to obtain the trained topic learning model.

[0015] In one embodiment, obtaining visual information corresponding to the first theme includes:

[0016] Inputting the first theme into the theme learning model in the theme library to obtain visual information corresponding to the first theme; or

[0017] The visual information corresponding to the first theme is obtained from the visual information corresponding to each of the themes stored in the theme library, wherein the visual information corresponding to each of the themes is obtained by inputting each of the themes into the theme learning model in advance.

[0018] In one embodiment, the shooting description information is text, and searching for a first topic that semantically matches the shooting description information includes:

[0019] Preprocessing the text to obtain a processed text;

[0020] Inputting the processed text into a BERT model to obtain text semantics;

[0021] Get the semantics of each topic;

[0022] The first topic is determined from among the topics according to the semantic matching degree between the semantics of each topic and the semantics of the text.

[0023] In one embodiment, outputting shooting guidance information corresponding to the first subject according to the visual information includes:

[0024] Outputting the visual information as shooting example information; and / or

[0025] The shooting guide information is generated and outputted according to the visual information.

[0026] In one embodiment, the method further comprises:

[0027] Acquire multiple images of the captured second subject;

[0028] Inputting the plurality of pictures into a typesetting model, and obtaining a typesetting position of each of the pictures determined by the typesetting model according to the content of each of the pictures;

[0029] According to the layout position of each of the pictures, each of the pictures is automatically layouted to obtain a theme picture album of the second theme.

[0030] In one embodiment, the training step of the typesetting model at least includes:

[0031] Obtaining a typeset sample of a theme picture album; the theme picture album sample includes various picture samples of the same theme;

[0032] Inputting each of the picture samples into an initial layout model to obtain a predicted layout position of each of the picture samples determined by the initial layout model according to the content of each of the picture samples;

[0033] The initial layout model is trained according to the predicted layout positions of the various picture samples and the actual layout positions of the various picture samples in the theme picture album samples to obtain the trained layout model.

[0034] In a second aspect, an embodiment of the present application provides a device for processing shooting guidance information, including:

[0035] A first acquisition module is used to acquire shooting description information;

[0036] A search module, used for searching for a first subject that semantically matches the shooting description information;

[0037] A second acquisition module is used to acquire visual information corresponding to the first theme; and

[0038] An output module is used to output shooting guidance information corresponding to the first theme based on the visual information.

[0039] In one embodiment, the second acquisition module includes:

[0040] An input unit is used to input the first topic into a topic learning model to obtain the visual information.

[0041] In one embodiment, the training step of the topic learning model at least includes:

[0042] Obtaining image samples of each subject sample, wherein the image samples carry visual annotation information;

[0043] Inputting the picture samples of each theme sample into the initial theme learning model to obtain the visual prediction information corresponding to each theme sample output by the initial theme learning model;

[0044] The initial topic learning model is trained according to the difference between the visual prediction information and the visual annotation information corresponding to each topic sample to obtain the trained topic learning model.

[0045] In one embodiment, the second acquisition module includes:

[0046] a theme input unit, used to input the first theme into the theme learning model in the theme library to obtain visual information corresponding to the first theme; or

[0047] The information acquisition unit is used to acquire visual information corresponding to the first theme from the visual information corresponding to each of the themes stored in the theme library, wherein the visual information corresponding to each of the themes is obtained by inputting each of the themes into the theme learning model in advance.

[0048] In one embodiment, the shooting description information is text, and the search module includes:

[0049] A preprocessing unit, used for preprocessing the text to obtain a processed text;

[0050] A text semantics acquisition unit, used to input the processed text into a BERT model to obtain text semantics;

[0051] A topic semantics acquisition unit, used to acquire the semantics of each topic;

[0052] A topic determination unit is used to determine the first topic from each of the topics according to the semantic matching degree between the semantics of each of the topics and the semantics of the text.

[0053] In one embodiment, the output module includes:

[0054] A first output unit, configured to output the visual information as shooting example information; and / or

[0055] The second output unit is used to generate and output the shooting guidance information according to the visual information.

[0056] In one embodiment, the device further comprises:

[0057] A picture acquisition module, used to acquire multiple pictures of the second subject;

[0058] A position determination module, used for inputting the plurality of pictures into a typesetting model, and obtaining a typesetting position of each of the pictures determined by the typesetting model according to the content of each of the pictures;

[0059] The automatic typesetting module is used to automatically typeset each of the pictures according to the typesetting position of each of the pictures to obtain a theme picture album of the second theme.

[0060] In one embodiment, the training step of the typesetting model at least includes:

[0061] Obtaining a typeset sample of a theme picture album; the theme picture album sample includes various picture samples of the same theme;

[0062] Inputting each of the picture samples into an initial layout model to obtain a predicted layout position of each of the picture samples determined by the initial layout model according to the content of each of the picture samples;

[0063] The initial layout model is trained according to the predicted layout positions of the various picture samples and the actual layout positions of the various picture samples in the theme picture album samples to obtain the trained layout model.

[0064] In a third aspect, an embodiment of the present application further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps in the above-mentioned method for processing shooting guidance information when executed by the processor.

[0065] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps in the above-mentioned method for processing shooting guidance information are implemented.

[0066] In a fifth aspect, the embodiments of the present application further provide a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. The processor of the computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the method provided in various optional implementations described in the embodiments of the present application.

[0067] In summary, in an embodiment of the present application, based on the shooting description information, a first theme that semantically matches the shooting description information can be determined, and then visual information corresponding to the first theme can be obtained, and shooting guidance information corresponding to the first theme can be output based on the visual information. In this way, the shooting guidance information can help users improve their shooting skills and take high-quality photos. And because the shooting guidance information is information corresponding to the first theme, the user can take theme photos based on the shooting guidance information. In addition, the first theme is a theme that semantically matches the shooting description information, so when the shooting description information is information input by the user, the shooting guidance information is the shooting guidance information of the theme that the user wants. In this way, the user can be guided to take high-quality theme photos. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the technical solutions in the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0069] Figure 1 is a schematic diagram of the steps of a method for processing shooting guidance information provided by an embodiment of the present application;

[0070] Figure 2 is a schematic diagram of steps of a method for determining a first subject provided by an embodiment of the present application;

[0071] Figure 3 It is a schematic diagram of the steps of a training method for a subject learning model provided in one embodiment of the present application;

[0072] Figure 4 is a schematic diagram of the steps of a method for processing shooting guidance information provided by an embodiment of the present application;

[0073] Figure 5 It is a schematic diagram of the steps of a method for training a typesetting model provided in an embodiment of the present application;

[0074] Figure 6 is a flowchart of a method for processing shooting guidance information provided by an embodiment of the present application;

[0075] Figure 7 This is a structural diagram of a device for processing shooting guidance information provided by an embodiment of the present application.

[0076] Figure 8 It is a schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in this application to clearly and completely describe the technical solutions in this application. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments in the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0078] In one embodiment, Figure 1 As shown, a method for processing shooting guide information is provided. Although a logical order is shown in the step diagram, in some cases, the steps shown or described may be performed in an order different from that shown in the drawings.

[0079] In one embodiment, the execution subject of the method for processing the shooting guidance information is an electronic device with a shooting function, and the electronic device can generate the shooting guidance information and directly display the shooting guidance information on a shooting page.

[0080] In another embodiment, the executor of the method for processing the shooting guidance information is an electronic device that does not have a shooting function. The electronic device generates and outputs the shooting guidance information, and then sends the shooting guidance information to an electronic device with a shooting function, and displays the shooting guidance information on the shooting page of the electronic device with a shooting function.

[0081] In summary, the method for processing shooting guidance information can be applied to an electronic device, wherein the electronic device may include a terminal or a server. The terminal may include, but is not limited to, one or more of a smart camera, a smart phone, a tablet computer, a portable computer, a desktop computer, and a car computer. The server may be a physical server or a cloud server that provides various cloud services. It is worth noting that this application does not limit the number of terminals or servers. According to implementation needs, there may be any number of terminals or servers. For example, the server may be a single server or a server cluster consisting of multiple servers, and so on.

[0082] The shooting guidance information of the embodiment of the present application can not only be used to guide users to take photos, but also be used to guide users to take dynamic photos and shoot videos.

[0083] The following is a detailed description of the method for processing shooting guidance information by taking a smartphone with a camera function as an example. It should be noted that the order of description of the following embodiments is not intended to limit the priority order of the embodiments.

[0084] according to Figure 1 The method for processing shooting guidance information shown in the figure at least includes steps S110 to S140, which are described in detail as follows:

[0085] In step S110 , the shooting description information is obtained.

[0086] The shot description is used to describe the subject of the image you took. The subject of an image is the core message or narrative that the image conveys. It refers to the main concept, object, or event that the image expresses. The subject of an image can be abstract, such as an emotion or idea, or concrete, such as a specific person or event.

[0087] The shooting description information may be information input by the user, or information provided to the user for the user to select. The shooting description information may be information in various modes such as text, image or video.

[0088] In step S120, a first subject that semantically matches the shooting description information is searched.

[0089] A plurality of topics are pre-constructed, and a first topic that matches the semantics of the shooting description information is searched from the pre-constructed plurality of topics. Matching the semantics of the shooting description information means that the distance between the semantics of the topic and the semantics of the shooting description information satisfies a matching condition; the matching condition may be the closest distance or the distance is less than a distance threshold, etc. The distance between the semantics may be a cosine distance, a Euclidean distance, and / or a Manhattan distance.

[0090] In one embodiment, the topic may be represented by text, and the semantics of each topic may be obtained through a natural language processing model or a pre-trained language representation model (Bidirectional Encoder Representations from Transformers, BERT model).

[0091] The semantics of the shooting description information is obtained through the processing model of the shooting description information. For shooting description information of different modalities, the processing model of the shooting description information can be different models, and the construction and training methods of the processing model of the shooting description information can refer to the relevant technology. For example, when the shooting description information is text, the processing model of the shooting description information can be a BERT model; when the shooting description information is an image, the processing model of the shooting description information can be a convolutional neural network; when the shooting description information is a video, the processing model of the shooting description information can be a recurrent neural network or a 3D convolutional neural network.

[0092] In one embodiment, the shooting description information is text, such as Figure 2 As shown, the step S120 may include steps S121 to S124.

[0093] In step S121, the text is preprocessed to obtain a processed text.

[0094] The shooting description information can be a continuous text or a text composed of multiple keywords. Preprocessing the text can convert the text into a format suitable for machine learning models to understand, which can improve the accuracy and efficiency of the subsequent text semantics.

[0095] The preprocessing includes, but is not limited to: text segmentation, part-of-speech tagging, stop word removal and / or text cleaning. Text segmentation is used to break down text into individual words or tokens, reducing words to their basic form, which can help the model recognize different word forms with the same root. Part-of-speech tagging assigns a part-of-speech label to each word in a sentence, which can improve the performance of the model. Stop words are words that do not provide much semantic information, and stop word removal can reduce the dimensionality of the model and improve computational efficiency. Text cleaning includes removing punctuation, converting uppercase and lowercase letters, and standardizing spelling, which can help the model focus on the semantic content of the text.

[0096] In step S122, the processed text is input into the BERT model to obtain text semantics.

[0097] The processed text is input into the pre-trained BERT model. The BERT model can perform semantic analysis on the processed text, extract the semantic information of words, determine the relationship between words, and output the text semantics of the text. The BERT model can have an in-depth semantic understanding of the text and has strong generalization ability.

[0098] In step S123, the semantics of each topic is obtained.

[0099] The method for obtaining the semantics of a topic can refer to the method for obtaining the text semantics of a text.

[0100] The semantics of the topics may be pre-acquired, and when searching for the first topic, the search is performed directly based on the semantics of the pre-acquired topics.

[0101] In step S124, the first topic is determined from the topics according to the semantic matching degree between the semantics of the topics and the semantics of the text.

[0102] The distance between the semantics of each topic and the semantics of the text is calculated, and the distance between the semantics of the topic and the semantics of the text is determined as the semantic matching degree of the semantics of the topic and the semantics of the text. Among the topics, the topic with the closest semantic distance is determined as the first topic.

[0103] In this way, by preprocessing the text, the accuracy of the text semantics can be improved and the efficiency can be improved; by matching the semantics to determine the first theme, it can be ensured that the determined first theme is the theme represented by the shooting description information that the user wants.

[0104] In step S130, visual information corresponding to the first theme is obtained.

[0105] Visual information includes visual feature information, sample images and / or environmental information. Visual feature information includes, but is not limited to, any one or more of the following: scene, composition, light, style and color. The sample images corresponding to a theme refer to typical images representing the theme. Environmental information includes: scene, props and / or costumes, etc.

[0106] In one embodiment, obtaining the visual information corresponding to the first theme may be inputting the first theme into a theme learning model to obtain the visual information. The theme learning model may automatically output the visual information corresponding to the theme according to the input theme. The training method of the theme learning model will be described in detail later.

[0107] In this way, based on the topic learning model, the visual information corresponding to the first topic can be quickly obtained, and then shooting guidance can be quickly provided to the user.

[0108] In step S140, shooting guidance information corresponding to the first subject is output according to the visual information.

[0109] The shooting guidance information can be used to guide the user to take a photo close to the visual information. Therefore, the shooting guidance information can be generated according to the visual information.

[0110] Under different themes, the photo that the shooting guidance information guides the user to take can be one photo or a series of photos. For example, when the theme is ID photo, the shooting guidance information is used to guide the user to take an ID photo; when the theme is wedding record, the shooting guidance information is used to guide the user to take a series of wedding-related photos. It is understandable that in the series of photos that the shooting guidance information guides the user to take, the visual information corresponding to each photo can have a focus.

[0111] In one embodiment, outputting the shooting guidance information corresponding to the first theme may include: outputting the visual information as shooting example information; and / or generating and outputting the shooting guidance information based on the visual information.

[0112] In one example, outputting the visual information as shooting example information may be directly displaying the visual information on the shooting page. For example, if the style represented by the visual information is dark, the shooting example information "style dark" may be directly displayed on the shooting page. When the visual information is an example picture, the example picture may be directly displayed on the shooting page.

[0113] In another example, generating and outputting the shooting guidance information according to the visual information may be generating guiding information according to the visual information and outputting the guiding information as the shooting guidance information. For example, if the light represented by the visual information is dim light, "pull down the small sun" may be displayed on the shooting page, where the small sun refers to a control on the shooting page for adjusting the light.

[0114] In another example, when the user is on a shooting page, visual information of the shooting page can be obtained in real time, and the visual information of the shooting page can be compared with the visual information of the first subject. Based on the difference between the visual information of the shooting page and the visual information of the first subject, shooting guidance information can be output to guide the user to shoot a subject photo with a smaller difference. It should be noted that obtaining the visual information of the shooting page requires the user's permission or consent, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0115] It is understandable that in order to provide more intuitive shooting guidance, the shooting guidance information is displayed on the shooting page. In one embodiment, the shooting guidance information can also be displayed on other pages, and the user can view the shooting guidance information on other pages and take photos on the shooting page. In one embodiment, the shooting guidance information can be displayed as a pop-up window independent of the shooting page at the same time as the shooting page.

[0116] In this way, through the shooting example information and / or the shooting guidance information, the user can be guided to shoot a theme photo that meets the visual information of the first theme. Moreover, the shooting guidance information guides the user to shoot a photo that not only meets the first theme, but also is a high-quality photo with aesthetic appeal.

[0117] By adopting the technical solution of the embodiment of the present application, based on the shooting description information, the first theme that semantically matches the shooting description information can be determined, and then the visual information corresponding to the first theme can be obtained, and the shooting guidance information corresponding to the first theme can be output based on the visual information. In this way, the shooting guidance information can help users improve their shooting skills and take high-quality photos. And because the shooting guidance information is information corresponding to the first theme, the user can take theme photos based on the shooting guidance information. In addition, the first theme is a theme that semantically matches the shooting description information. Therefore, when the shooting description information is information input by the user, the shooting guidance information is the shooting guidance information of the theme that the user wants. In this way, the user can be guided to take high-quality theme photos.

[0118] Based on the above technical solution, as an embodiment, Figure 3 As shown, the training steps of the topic learning model at least include steps S210 to S230.

[0119] In step S210, image samples of each theme sample are obtained, where the image samples carry visual annotation information.

[0120] A large number of image samples are collected, and theme samples of the image samples are determined, and the image samples are annotated to obtain visual annotation information of the image samples. In one embodiment, the theme samples may include: urban landscapes, portraits, and animal photography. The visual annotation information may include fresh style, retro style, and dark style.

[0121] In one embodiment, the image samples may be preprocessed, including unifying the size and format of the images.

[0122] In step S220, the picture samples of each theme sample are input into the initial theme learning model to obtain the visual prediction information corresponding to each theme sample output by the initial theme learning model.

[0123] The initial topic learning model is a topic learning model to be trained. The image samples of each topic sample are input into the initial topic learning model, and the initial topic learning model can output visual prediction information according to each image sample.

[0124] In step S230, the initial topic learning model is trained according to the difference between the visual prediction information and the visual annotation information corresponding to each topic sample to obtain the trained topic learning model.

[0125] According to the difference between the visual prediction information and the visual annotation information corresponding to each topic sample, a loss function is established, and the initial topic learning model is trained with minimizing the loss function as the training goal until the model converges or the training number is reached, thereby obtaining a trained topic learning model.

[0126] By adopting the technical solution of the embodiment of the present application, the initial topic learning model is trained to learn the visual information corresponding to each topic, so that the trained topic learning model can quickly determine the visual information corresponding to the input topic.

[0127] On the basis of the above technical solution, a theme library can be constructed based on a trained theme learning model. The obtaining of the visual information corresponding to the first theme may include: inputting the first theme into the theme learning model in the theme library to obtain the visual information corresponding to the first theme; or obtaining the visual information corresponding to the first theme from the visual information corresponding to each theme stored in the theme library, wherein the visual information corresponding to each theme is obtained by inputting each theme into the theme learning model in advance.

[0128] In one embodiment, the theme library includes a theme learning model and various themes. After determining a first theme from the theme library based on the shooting description information, the first theme is input into the theme learning model to obtain visual information corresponding to the first theme.

[0129] In this way, the amount of data stored in the theme library is less, which can save storage resources.

[0130] In another embodiment, the theme library includes each theme and visual information corresponding to each theme. The visual information corresponding to each theme is obtained by inputting each theme into the theme learning model in advance, and then the visual information corresponding to each theme and each theme are stored in the theme library. After determining the first theme from the theme library, the visual information corresponding to the first theme can be directly obtained from the theme library.

[0131] In this way, the visual information corresponding to the first theme is directly obtained from the theme library without the need to use the theme learning model for online calculation, which can improve the efficiency of outputting the shooting guidance information.

[0132] Based on the above technical solution, Figure 4 As shown, the method for processing shooting guide information may further include steps S310 to S330.

[0133] In step S310, a plurality of photographed pictures of the second subject are acquired.

[0134] The second theme may be the same as or different from the first theme.

[0135] In one embodiment, the second theme is the same as the first theme, and shooting is performed based on shooting guidance information corresponding to the first theme to obtain multiple pictures of the first theme (ie, the second theme), and then the multiple pictures of the first theme are arranged.

[0136] In another embodiment, the second theme is different from the first theme, and the multiple pictures of the second theme may be pictures taken by the user without receiving shooting guidance. The multiple pictures of the second theme may be directly obtained, and then the multiple pictures of the second theme may be typeset.

[0137] In step S320, the multiple pictures are input into a typesetting model to obtain a typesetting position of each picture determined by the typesetting model according to the content of each picture.

[0138] The layout model learns the layout positions of images of different contents under different topics. The training method of the layout model will be described in detail later.

[0139] In one embodiment, multiple pictures are input into a typesetting model, which can automatically identify the second theme corresponding to the multiple pictures and the content of each picture, and then determine the layout position of each picture under the second theme based on the content of each picture.

[0140] In another embodiment, multiple pictures and the second theme can be directly input into the typesetting model, the typesetting model directly identifies the content of each picture, and then determines the typesetting position of each picture under the second theme based on the content of each picture.

[0141] In step S330, each of the pictures is automatically laid out according to the layout position of each of the pictures to obtain a theme picture album of the second theme.

[0142] According to the determined layout positions of the various pictures, the various pictures are automatically layouted at the corresponding layout positions to obtain a theme picture album of the second theme.

[0143] As an embodiment, when a theme picture album is generated, patterns, frames, etc. corresponding to the theme can also be automatically added to the theme picture album.

[0144] By adopting the technical solution of the embodiment of the present application, multiple pictures of a theme are automatically laid out, which can improve the aesthetics of the theme picture album and save user time.

[0145] Based on the above technical solution, as an embodiment, Figure 5 As shown, the training steps of the typesetting model at least include steps S410 to S430.

[0146] In step S410, a typeset sample of a theme picture album is obtained; the theme picture album sample includes various picture samples of the same theme.

[0147] Each theme picture album sample includes picture samples with the same theme. The theme picture album samples may be manually typeset.

[0148] In step S420, each of the picture samples is input into an initial layout model to obtain a predicted layout position of each of the picture samples determined by the initial layout model according to the content of each of the picture samples.

[0149] The initial layout model is the layout model to be trained.

[0150] In one embodiment, each image sample is input into an initial typesetting model, which automatically identifies the topics corresponding to the multiple image samples and the content of each image sample, and then predicts the predicted typesetting position of each image sample under the topic based on the content of each image sample.

[0151] In another embodiment, each picture sample and the theme of each picture sample are input into an initial typesetting model, the initial typesetting model directly identifies the content of each picture sample, and then predicts the predicted typesetting position of each picture sample under the theme based on the content of each picture sample.

[0152] In step S430, the initial layout model is trained according to the predicted layout positions of the various picture samples and the actual layout positions of the various picture samples in the theme picture album samples to obtain the trained layout model.

[0153] According to the predicted layout position of each picture sample and the actual layout position of each picture sample in the theme picture album sample, a position loss function is established. The initial layout model is trained with the goal of minimizing the position loss function until the model converges or the training number is reached to obtain a trained layout model.

[0154] By adopting the technical solution of the embodiment of the present application, the initial typesetting model is trained to learn the typesetting positions corresponding to each image sample under each theme, so that the trained typesetting model can quickly typeset multiple images for multiple images of the same input theme, which helps to save user time.

[0155] Based on the above technical solution, as an embodiment, Figure 6 As shown in the figure, the user inputs the shooting description information: lawn wedding. The BERT model can obtain the text semantics of the shooting description information. By semantically matching the semantics of the shooting description information and the semantics of the subject, it can be determined that the first subject is an outdoor wedding. The number of photos to be taken is determined according to the first subject, and when taking each photo, the scene, composition suggestions and other shooting guidance information of each photo are generated according to the visual information of the outdoor wedding. After the photo is taken, a series of theme photos of the outdoor wedding can be obtained, and these theme photos can be automatically typeset to obtain a theme photo album of the outdoor wedding.

[0156] By adopting the technical solution of the embodiment of the present application, the shooting guidance information can help users improve their shooting skills, take high-quality photos of the theme that the user wants, and can layout the theme photos to obtain a theme photo album, which can effectively improve the user experience and save user time.

[0157] In order to better implement the method for processing shooting guidance information of the present application, the present application also provides a device for processing shooting guidance information based on the above-mentioned method for processing shooting guidance information. The meanings of the terms are the same as those in the above-mentioned method for processing shooting guidance information, and the specific implementation details can refer to the description in the method embodiment.

[0158] See also Figure 7 , Figure 7 : is a structural diagram of a device for processing shooting guidance information provided in an embodiment of the present application, the device for processing shooting guidance information includes:

[0159] The first acquisition module 701 is used to acquire shooting description information;

[0160] A search module 702, configured to search for a first subject that semantically matches the shooting description information;

[0161] A second acquisition module 703 is used to acquire visual information corresponding to the first theme; and

[0162] The output module 704 is used to output shooting guidance information corresponding to the first theme according to the visual information.

[0163] In one embodiment, the second acquisition module 703 includes:

[0164] An input unit is used to input the first topic into a topic learning model to obtain the visual information.

[0165] In one embodiment, the training step of the topic learning model at least includes:

[0166] Obtaining image samples of each subject sample, wherein the image samples carry visual annotation information;

[0167] Inputting the picture samples of each theme sample into the initial theme learning model to obtain the visual prediction information corresponding to each theme sample output by the initial theme learning model;

[0168] The initial topic learning model is trained according to the difference between the visual prediction information and the visual annotation information corresponding to each topic sample to obtain the trained topic learning model.

[0169] In one embodiment, the second acquisition module 703 includes:

[0170] a theme input unit, used to input the first theme into the theme learning model in the theme library to obtain visual information corresponding to the first theme; or

[0171] The information acquisition unit is used to acquire visual information corresponding to the first theme from the visual information corresponding to each of the themes stored in the theme library, wherein the visual information corresponding to each of the themes is obtained by inputting each of the themes into the theme learning model in advance.

[0172] In one embodiment, the shooting description information is text, and the search module 702 includes:

[0173] A preprocessing unit, used for preprocessing the text to obtain a processed text;

[0174] A text semantics acquisition unit, used to input the processed text into a BERT model to obtain text semantics;

[0175] A topic semantics acquisition unit, used to acquire the semantics of each topic;

[0176] A topic determination unit is used to determine the first topic from each of the topics according to the semantic matching degree between the semantics of each of the topics and the semantics of the text.

[0177] In one embodiment, the output module 704 includes:

[0178] A first output unit, configured to output the visual information as shooting example information; and / or

[0179] The second output unit is used to generate and output the shooting guidance information according to the visual information.

[0180] In one embodiment, the device further comprises:

[0181] A picture acquisition module, used to acquire multiple pictures of the second subject;

[0182] A position determination module, used for inputting the plurality of pictures into a typesetting model, and obtaining a typesetting position of each of the pictures determined by the typesetting model according to the content of each of the pictures;

[0183] The automatic typesetting module is used to automatically typeset each of the pictures according to the typesetting position of each of the pictures to obtain a theme picture album of the second theme.

[0184] In one embodiment, the training step of the typesetting model at least includes:

[0185] Obtaining a typeset sample of a theme picture album; the theme picture album sample includes various picture samples of the same theme;

[0186] Inputting each of the picture samples into an initial layout model to obtain a predicted layout position of each of the picture samples determined by the initial layout model according to the content of each of the picture samples;

[0187] The initial layout model is trained according to the predicted layout positions of the various picture samples and the actual layout positions of the various picture samples in the theme picture album samples to obtain the trained layout model.

[0188] By adopting the technical solution of the embodiment of the present application, based on the shooting description information, the first theme that semantically matches the shooting description information can be determined, and then the visual information corresponding to the first theme can be obtained, and the shooting guidance information corresponding to the first theme can be output based on the visual information. In this way, the shooting guidance information can help users improve their shooting skills and take high-quality photos. And because the shooting guidance information is information corresponding to the first theme, the user can take theme photos based on the shooting guidance information. In addition, the first theme is a theme that semantically matches the shooting description information. Therefore, when the shooting description information is information input by the user, the shooting guidance information is the shooting guidance information of the theme that the user wants. In this way, the user can be guided to take high-quality theme photos.

[0189] For the specific definition of the processing device of the shooting guidance information, please refer to the definition of the processing method of the shooting guidance information above, which will not be repeated here. Each module in the above-mentioned processing device of the shooting guidance information can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0190] In addition, the present application also provides an electronic device, such as Figure 8 As shown, it shows a schematic diagram of the structure of the electronic device involved in this application, specifically:

[0191] The electronic device may include one or more processors 801 of processing cores and one or more computer-readable storage media memories 802 and other components. Those skilled in the art will appreciate that Figure 8 The electronic device structure shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0192] The processor 801 is the control center of the electronic device. It uses various interfaces and lines to connect various parts of the entire electronic device. By running or executing software programs and / or modules stored in the memory 802 and calling data stored in the memory 802, it performs various functions of the electronic device and processes data, thereby monitoring the electronic device as a whole. Optionally, the processor 801 may include one or more processing cores; preferably, the processor 801 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 801.

[0193] The memory 802 can be used to store software programs and modules. The processor 801 executes various functional applications and data processing by running the software programs and modules stored in the memory 802. The memory 802 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 802 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory 802 may also include a memory controller to provide the processor 801 with access to the memory 802.

[0194] In one embodiment, the electronic device further includes a power supply 803 for supplying power to various components. Preferably, the power supply 803 can be logically connected to the processor 801 through a power management system, so as to manage charging, discharging, power consumption and other functions through the power management system. The power supply 803 can also include one or more DC or AC power supplies, recharging systems, power supply device debugging circuits, power converters or inverters, power status indicators and other arbitrary components.

[0195] In one embodiment, the electronic device may further include an input unit 804, which may be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control.

[0196] Although not shown, the electronic device may further include a display unit, etc., which will not be described in detail herein. Specifically in this embodiment, the processor 801 in the electronic device will load the executable files corresponding to the processes of one or more application programs into the memory 802 according to the following instructions, and the processor 801 will run the application programs stored in the memory 802, thereby implementing the steps in any one of the methods for processing shooting guidance information provided in the embodiments of the present application.

[0197] Those skilled in the art will understand that Figure 8 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0198] In one embodiment, an electronic device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the method described in any embodiment of the present application is implemented.

[0199] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method described in any embodiment of the present application is implemented.

[0200] In some embodiments, a computer program product is also proposed, including a computer program or instructions, which implement the method described in any embodiment of the present application when executed by a processor.

[0201] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0202] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0203] To this end, the present application provides a computer-readable storage medium, on which a computer program is stored. The computer program can be loaded by a processor to execute the steps in any method for processing shooting guidance information provided in the present application.

[0204] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0205] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0206] Since the instructions stored in the computer-readable storage medium can execute the steps in any method for processing shooting guidance information provided in the present application, the beneficial effects that can be achieved by any method for processing shooting guidance information provided in the present application can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0207] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device that includes a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also includes elements that are inherent to such process, method, article or terminal device.

[0208] The above is a detailed introduction to a method, device, electronic device and computer-readable storage medium for processing shooting guidance information provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for processing shooting guidance information, characterized in that: include: Get the shooting description information; Finding a first subject that semantically matches the shooting description information; Acquire visual information corresponding to the first theme; as well as Based on the visual information, shooting guidance information corresponding to the first theme is output.

2. The method according to claim 1, characterized in that The obtaining of visual information corresponding to the first theme includes: The first topic is input into a topic learning model to obtain the visual information.

3. The method according to claim 2, characterized in that The training steps of the topic learning model at least include: Obtaining image samples of each subject sample, wherein the image samples carry visual annotation information; Inputting the picture samples of each theme sample into the initial theme learning model to obtain the visual prediction information corresponding to each theme sample output by the initial theme learning model; The initial topic learning model is trained according to the difference between the visual prediction information and the visual annotation information corresponding to each topic sample to obtain the trained topic learning model.

4. The method according to claim 2, characterized in that: The obtaining of visual information corresponding to the first theme includes: Inputting the first theme into the theme learning model in the theme library to obtain visual information corresponding to the first theme; or The visual information corresponding to the first theme is obtained from the visual information corresponding to each of the themes stored in the theme library, wherein the visual information corresponding to each of the themes is obtained by inputting each of the themes into the theme learning model in advance.

5. The method according to any one of claims 1 to 4, characterized in that: The shooting description information is text, and searching for a first topic that semantically matches the shooting description information includes: Preprocessing the text to obtain a processed text; Inputting the processed text into a BERT model to obtain text semantics; Get the semantics of each topic; The first topic is determined from among the topics according to the semantic matching degree between the semantics of each topic and the semantics of the text.

6. The method according to claim 1, characterized in that Outputting shooting guidance information corresponding to the first subject according to the visual information includes: Outputting the visual information as shooting example information; and / or The shooting guide information is generated and outputted according to the visual information.

7. The method according to claim 1, characterized in that The method further comprises: Acquire multiple images of the captured second subject; Inputting the plurality of pictures into a typesetting model, and obtaining a typesetting position of each of the pictures determined by the typesetting model according to the content of each of the pictures; According to the layout position of each of the pictures, each of the pictures is automatically layouted to obtain a theme picture album of the second theme.

8. The method according to claim 7, characterized in that The training steps of the typesetting model at least include: Obtaining a typeset sample of a theme picture album; the theme picture album sample includes various picture samples of the same theme; Inputting each of the picture samples into an initial layout model to obtain a predicted layout position of each of the picture samples determined by the initial layout model according to the content of each of the picture samples; The initial layout model is trained according to the predicted layout positions of the various picture samples and the actual layout positions of the various picture samples in the theme picture album samples to obtain the trained layout model.

9. A device for processing shooting guidance information, characterized in that: include: A first acquisition module is used to acquire shooting description information; A search module, used for searching for a first subject that semantically matches the shooting description information; A second acquisition module, used to acquire visual information corresponding to the first theme; as well as An output module is used to output shooting guidance information corresponding to the first theme based on the visual information.

10. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps in the method for processing shooting guide information as described in any one of claims 1 to 8 when executing the computer program.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the method for processing shooting guide information according to any one of claims 1 to 8 are implemented.

Images