Material generation method and device

Abstract

The embodiment of the invention provides a material generation method and device, computer equipment, a computer readable storage medium and a computer program product, and belongs to the technical field of computers. The method comprises the steps of obtaining an initial material; obtaining an intermediate material including a target element based on the initial material through a pre-trained material processing model; taking the intermediate material and the copywriting information for the intermediate material as model input, and generating a target material through a pre-trained material generation model; wherein the target material comprises the target element and the copywriting information. According to the technical scheme provided by the embodiment of the invention, the intelligence and automation level of material generation can be improved, the material production time is saved, the material generation efficiency is improved, and meanwhile, the material production threshold is reduced.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of computer technologies, and in particular, to a method and apparatus for generating materials, a computer device, a computer-readable storage medium, and a computer program product. Background Art

[0002] In the process of product promotion, the use of promotional materials such as promotional media (pictures, videos, etc.) is particularly important. A good promotional material can quickly attract the attention of the target audience, make them interested in the promoted item, and thus effectively improve the market awareness and sales conversion rate of the product. However, when producing promotional materials, the degree of automation and intelligence of the promoter is low, and a large amount of time and resources need to be invested, resulting in low efficiency and quality of material production.

[0003] It should be noted that the above content is not necessarily prior art and is not used to limit the scope of patent protection of the present application. Summary of the Invention

[0004] Embodiments of the present application provide a method and apparatus for generating materials, a computer device, a computer-readable storage medium, and a computer program product to solve or alleviate one or more of the above technical problems.

[0005] One aspect of the embodiments of the present application provides a method for generating materials, the method including: Obtaining an initial material; Based on the initial material, obtaining an intermediate material including target elements through a pre-trained material processing model; Using the intermediate material and the copywriting information for the intermediate material as model inputs, and generating a target material through a pre-trained material generation model; wherein, the target material includes the target elements and the copywriting information.

[0006] Optionally, obtaining an intermediate material including target elements based on the initial material includes: Determining the material quality of the initial material; In the case where the material quality of the initial material is lower than a preset standard, processing the initial material to obtain an optimized material; Using the optimized material as a model input, and generating the intermediate material through the material processing model according to the optimized material.

[0007] Optionally, using the intermediate material and the copywriting information for the intermediate material as model inputs, and generating a target material through a pre-trained material generation model includes: Generating a material background based on the copywriting information and adapted to the intermediate material through the material generation model; Combine the intermediate material to the material adaptation position of the material background to obtain the target material.

[0008] Optionally, the method further includes: Input the target material into a pre-trained material recognition model to obtain the recognition result of the target material; wherein, the material recognition model is used to determine whether there is any illegal content in the target material; In the case that the target material has illegal content, use the intermediate material and the copywriting information as model inputs, and generate a first remade material through the material generation model, and the first remade material is different from the target material.

[0009] Optionally, the method further includes: Input the target material into a pre-trained click-through rate prediction model to output the predicted click-through rate of the target material through the click-through rate prediction model; In the case that the predicted click-through rate is lower than a preset value, use the intermediate material and the copywriting information as model inputs, and generate a second remade material through the material generation model, and the second remade material is different from the target material.

[0010] Optionally, the method further includes: Return the target material to the target object; In response to the target object selecting the target material, save the target material and the corresponding intermediate material, copywriting information, and / or material background.

[0011] Another aspect of the embodiments of the present application provides a material generation device, and the device includes: A first acquisition module, configured to acquire an initial material; A second acquisition module, configured to obtain an intermediate material including target elements based on the initial material through a pre-trained material processing model; A generation module, configured to use the intermediate material and the copywriting information for the intermediate material as model inputs, and generate a target material through a pre-trained material generation model; wherein, the target material includes the target elements and the copywriting information.

[0012] Another aspect of the embodiments of the present application provides a computer device, including: At least one processor; and A memory communicatively connected to the at least one processor; Wherein: the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method as described above.

[0013] Another aspect of the embodiments of the present application provides a computer-readable storage medium, in which computer instructions are stored, and when the computer instructions are executed by a processor, the method described above is implemented.

[0014] Another aspect of the embodiments of the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0015] The embodiments of the present application adopting the above technical solutions may include the following advantages: The model generates target materials based on initial materials and copywriting information. Thereby, the intelligence and automation level of material generation can be improved, the production time of materials can be saved, the generation efficiency of materials is improved, and at the same time, the production threshold of materials is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings exemplarily show the embodiments and constitute a part of the description, and are used together with the written description of the description to explain the exemplary embodiments of the embodiments. The shown embodiments are only for illustrative purposes and do not limit the scope of the claims. In all the drawings, the same reference numerals refer to similar but not necessarily the same elements.

[0017] Figure 1 Schematically shows an operating environment diagram of the material generation method according to Embodiment 1 of the present application; Figure 2 Schematically shows a flowchart of the material generation method according to Embodiment 1 of the present application; Figure 3 Schematically shows Figure 2 a sub-step flowchart of step S202 in; Figure 4 Schematically shows Figure 2 a sub-step flowchart of step S204 in; Figure 5 Schematically shows an additional flowchart of the material generation method according to Embodiment 1 of the present application; Figure 6 Schematically shows another additional flowchart of the material generation method according to Embodiment 1 of the present application; Figure 7 Schematically shows yet another additional flowchart of the material generation method according to Embodiment 1 of the present application; Figure 8 Schematically shows an exemplary application flowchart of the material generation method according to Embodiment 1 of the present application; Figure 9 Schematically shows a block diagram of the material generation device according to Embodiment 2 of the present application; and Figure 10 Schematically shows a schematic diagram of the hardware architecture of a computer device according to Embodiment 3 of the present application. Detailed implementation manners

[0018] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0019] It should be noted that the descriptions involving "first", "second", etc. in the embodiments of the present application are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In addition, the technical solutions between various embodiments may be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement them. When the combination of technical solutions appears to be contradictory or unable to be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present application.

[0020] In the description of the present application, it should be understood that the numerical labels before the steps do not identify the order of execution of the steps, but are only used to facilitate the description of the present application and to distinguish each step, and thus cannot be understood as a limitation to the present application.

[0021] First, provide the term explanations involved in the present application: SKU (Stock Keeping Unit, inventory unit), which can be used to distinguish commodity units with different specifications, colors, sizes, configurations and other attributes in the same category of commodities.

[0022] SKU diagram, which is used to display the detailed information and features of commodities, and can include the display of various angles such as the front, side, back, and details of the commodities, so as to visually present the detailed information such as the appearance, color, and size of the commodities.

[0023] Convolutional Neural Network (CNN): A deep learning model used to process data with a grid structure, such as images, videos, etc.

[0024] Recurrent Neural Network (RNN): It is a neural network architecture used to process sequential data. By introducing a recurrent structure, the network can utilize the information from the previous moment to influence the output at the current moment, thereby capturing the temporal dependencies in sequential data.

[0025] Deep Neural Network (DNN): It is an artificial neural network with a multi-layer structure. By simulating the connection mode of human brain neurons, it uses a large number of hierarchical structures to handle complex pattern recognition and data fitting problems.

[0026] CTR (Click-Through Rate): It is used to measure the frequency of being clicked by users for advertisements, links, search results, etc. It is obtained by calculating the ratio of the number of clicks to the number of impressions, and the formula is CTR = (number of clicks / number of impressions) × 100%.

[0027] Conversion rate: It is a ratio that measures the number of users who complete specific target behaviors (such as purchasing goods, registering accounts, clicking links, etc.) within a certain period of time to the total number of visits or potential users.

[0028] Secondly, to facilitate the understanding of the technical solutions provided by the embodiments of the present application by those skilled in the art, the related technologies are described below: In the current process of placing commodity advertisements, the advertisement placement of most advertisers is limited by the material production ability and material production efficiency, resulting in a low efficiency in building advertisement plans. At the same time, due to problems such as the lack of professional design teams, insufficient visual creativity reserves, and high technical thresholds for multimedia production, many advertisers are difficult to produce advertisement materials with good promotion effects, which has a negative impact on the placement effects of advertisers.

[0029] Therefore, the embodiments of the present application provide a technical solution for material generation. In this technical solution, (1) AI (Artificial Intelligence) capabilities are integrated into the production of commodity category creative materials. When users upload any style of SKU images, AI can systematically optimize and perform matte extraction on the SKU images, and then produce corresponding commodity marketing materials according to the relevant requirement instructions input by users, reducing the production threshold of commodity category creative materials, improving the creative material production efficiency of advertisers, and thus enhancing the placement effects and revenues of advertisers; (2) Users can save the AI processing solutions for subsequent reuse of the solutions that meet the expectations. See the following text for details.

[0030] Finally, for the convenience of understanding, an exemplary operating environment is provided below.

[0031] As Figure 1 shown, the operating environment diagram includes: service platform 2, network 4, client 6, where: Server 2 can be composed of a single or multiple computing devices. The multiple computing devices may include virtualized computing instances. The virtualized computing instances may include virtual machines, such as emulations of computer systems, operating systems, servers, etc. The computing devices can load virtual machines based on virtual images and / or other data that define specific software (e.g., operating systems, dedicated applications, servers) for emulation. As the demand for different types of processing services changes, different virtual machines can be loaded and / or terminated on one or more computing devices. A hypervisor can be implemented to manage the use of different virtual machines on the same computing device.

[0032] Server 2 can be configured to communicate with client 6 etc. via network 4. Network 4 includes various network devices, such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices, and / or the like. Network 4 can include physical links, such as coaxial cable links, twisted pair cable links, fiber optic links and combinations thereof, etc., or wireless links, such as cellular links, satellite links, Wi-Fi links, etc.

[0033] Service platform 2 can provide services such as storage, reading, writing, querying, deleting, etc., such as providing a material storage service for clients.

[0034] Client 6 can be an electronic device running an operating system such as Windows, Android™, or iOS, such as a smart phone, tablet device, laptop computer, virtual reality device, gaming device, set-top box, in-vehicle terminal, smart TV. Based on the above operating systems, various application programs can be run, such as a material upload program.

[0035] Client 6 can provide / configure a user access page for manipulating service platform 2 or uploading objects, etc.

[0036] Note that the above devices are exemplary, and the number and types of devices can be adjusted in different scenarios or according to different requirements.

[0037] The following takes service platform 2 as the execution subject and introduces the technical solutions of this application through multiple embodiments. It should be noted that these embodiments can be implemented in various different forms and should not be construed as being limited only to the embodiments described herein.

[0038] Embodiment 1 Figure 2 A flowchart of a material generation method according to Embodiment 1 of the present application is schematically shown.

[0039] As Figure 2 shown, the material generation method may include steps S200 to S204, where: Step S200, obtain initial materials; In step S202, based on the initial material, an intermediate material including target elements is obtained through a pre-trained material processing model. In step S204, the intermediate material and the copywriting information for the intermediate material are used as model inputs, and a target material is generated through a pre-trained material generation model; wherein, the target material includes the target elements and the copywriting information.

[0040] The material generation method provided in this embodiment generates a target material through a model based on the initial material and the copywriting information. Thereby, the intelligence and automation level of material generation can be improved, the production time of materials can be saved, the generation efficiency of materials can be improved, and at the same time, the production threshold of materials can be reduced.

[0041] The following combines Figure 2 to elaborate in detail on each step in steps S200 to S204 and other optional steps.

[0042] Step S200 to obtain the initial material.

[0043] The initial material can be a picture, a video, an audio, etc. The obtained initial material can be selected from a default material library, or obtained from a material website, etc., or uploaded by a user through an upload interface (such as a web page, a mobile application, etc.).

[0044] Step S202 Based on the initial material, an intermediate material including target elements is obtained through a pre-trained material processing model.

[0045] The material processing model can be based on a third-party open-source machine learning model, a neural network model, etc., and can be trained through pre-annotated SKU map data. The SKU map data can include SKU maps of different products and different scenarios.

[0046] The target element can be the main element in the initial material, such as the product main body in the SKU picture, etc. According to actual needs, the target element can also include elements related to the main element, such as the bracket for supporting the product main body in the SKU picture.

[0047] For example, the initial material is a photo of a porcelain plate, which includes an upright porcelain plate, a wooden bracket for placing the porcelain plate, and a marble tabletop for placing the wooden bracket. Then the material processing model can extract the porcelain plate and the wooden bracket from it and output them as the intermediate material. After obtaining the intermediate material, it can also be saved to a fixed storage location (such as the SKU picture library) for subsequent repeated use.

[0048] Extract the target elements from the initial material through the model. Thereby, the interference of other elements irrelevant to the target elements in the initial material can be excluded, the information density of the obtained intermediate material can be improved, and subsequent material analysis and processing can be carried out more efficiently.

[0049] Before obtaining the intermediate material based on the initial material, some preprocessing operations can also be performed on the initial material. The following provides an exemplary preprocessing operation.

[0050] In an alternative embodiment, as Figure 3 shown, step S202 includes: S300, determining the material quality of the initial material.

[0051] S302, in the case where the material quality of the initial material is lower than the preset standard, processing the initial material to obtain an optimized material.

[0052] S304, using the optimized material as the model input, and generating the intermediate material from the optimized material through the material processing model.

[0053] The material quality may include the usability of the initial material, such as whether there are defects or obstructions in the commodity main body, whether there are problems such as poor clarity or poor shooting angle. It may also include the estimated effect of the initial material, such as the estimated CTR (click-through rate) of the promotional material generated from the initial material. The process of determining the material quality can be manual screening or detection using tools such as models.

[0054] The processing performed on the initial material may include intelligent patching, intelligent redrawing, or intelligent replacement, etc. Intelligent patching can repair the defective or obstructed material main body in the initial material. Intelligent redrawing can redraw the initial material with low clarity or poor shooting angle. Intelligent replacement can identify the material main body information in the initial material that is estimated to have a negative impact on the CTR, and obtain a new material containing the material main body through channels such as a material library to replace the initial material.

[0055] Optimizing the initial material with low quality before using it to obtain the intermediate material can improve the success rate of obtaining the intermediate material, and at the same time improve the quality and effect of the finally generated promotional material, enhancing the attractiveness and conversion rate of the promotional activity.

[0056] Step S204 , using the intermediate material and the copywriting information for the intermediate material as the model input, and generating the target material through a pre-trained material generation model; wherein, the target material includes the target elements and the copywriting information.

[0057] The copywriting information may include the name, price, or marketing selling points of the target element (such as the commodity main body). It may also include the quantity, size, and design style of the target material to be generated, etc.

[0058] The material generation model can be based on a deep learning model, etc., and is trained using public commodity pictures (i.e., intermediate materials) and marketing information (i.e., copywriting information) as training data.

[0059] For example, the intermediate material is a porcelain plate placed vertically on a wooden stand. The copywriting information includes the brand name of the porcelain plate "Blue and White Porcelain", the price "99 yuan", the marketing selling points "handmade, unique texture", and the generation requirements for the target material: the generation quantity is 1 piece, the size is 1920×1080 pixels, and the design style is minimalist. The finally generated target material has a pure white background, the porcelain plate in the picture is placed on the wooden stand, and there is a copywriting beside it "Blue and White Porcelain, 99 yuan, handmade, unique texture", the font is bold, the font size is moderate, the font color is black, and the overall layout is simple and beautiful.

[0060] By using the model to produce the corresponding target material according to the input intermediate material and copywriting information, the production threshold of the material can be reduced, and at the same time, the intelligence and automation level of material generation are improved, the material production efficiency is enhanced, and thus the promotion effect of the promoter is improved.

[0061] There are various ways to generate the target material according to the intermediate material and copywriting information. The following provides an exemplary way to generate the target material.

[0062] In an alternative embodiment, as Figure 4 shown, step S204 includes: S400, generating a material background based on the copywriting information and adapted to the intermediate material through the material generation model.

[0063] S402, combining the intermediate material to the material adaptation position of the material background to obtain the target material.

[0064] For example, the intermediate material is a porcelain plate placed vertically on a wooden stand, and the copywriting information mentions that the design style is minimalist. Then the model can generate a pure white background and indicate that the intermediate material should be placed in the center of the background.

[0065] Generate the material background according to the requirements and combine it with the intermediate material to obtain the target material. Thus, the harmony between the target element and the material background can be improved, thereby enhancing the visual effect and attractiveness of the target material, better meeting the visual needs of users for promotion materials, and further enhancing the promotion effect.

[0066] After generating the target material, the generated target material can also be evaluated, and when the evaluation fails, the target material can be remade. The following provides several exemplary evaluation points.

[0067] Evaluation point 1: As Figure 5 shown, the method further includes: S500, inputting the target material into a pre-trained material recognition model to obtain the recognition result of the target material; wherein, the material recognition model is used to determine whether there is any illegal content in the target material.

[0068] S502, in the case where there is illegal content in the target material, using the intermediate material and the copywriting information as model inputs, and generating a first remade material through the material generation model, the first remade material being different from the target material.

[0069] The material recognition model can be based on a convolutional neural network (CNN), a recurrent neural network (RNN), etc., and is trained by materials containing illegal content through manual annotation or automatic annotation. The illegal content can include sensitive content, false content, and inductive content, etc.

[0070] After using the material recognition model to identify the illegal content of the target material, it can also be transferred to a manual node for re-audit to further determine whether the target material is illegal. The material generation model can generate a first remade material that is different from the target material and meets the requirements of advertisement review by changing the content, style, layout, etc. of the copywriting.

[0071] When illegal content is identified in the target material, the target material is remade. Thus, it can be ensured that the generated material meets the review standards, avoiding the failure of promotion content delivery due to illegal content, thereby improving the safety and reliability of promotion delivery.

[0072] Evaluation point 2: As Figure 6 shown, the method further includes: S600, inputting the target material into a pre-trained click-through rate prediction model to output the predicted click-through rate of the target material through the click-through rate prediction model.

[0073] S602, in the case where the predicted click-through rate is lower than a preset value, using the intermediate material and the copywriting information as model inputs, and generating a second remade material through the material generation model, the second remade material being different from the target material.

[0074] The click-through rate prediction model can be based on a deep neural network (DNN), a convolutional neural network (CNN), or a recurrent neural network (RNN), etc., and is trained using historical advertising materials (including images, copywriting, click-through rates, etc.) as training data. In addition to the click-through rate, other metrics such as the conversion rate of the target material can also be predicted.

[0075] The material generation model can generate a second remade material that is different from the target material and has a higher predicted click-through rate by changing the content, style, layout, etc. of the copywriting. Before remaking, the user can also be confirmed whether to perform remaking.

[0076] Remake the target material with a low predicted click-through rate. Thus, the predicted click-through rate of the remade material can be improved, thereby enhancing the promotion effect of the remade material, helping users improve the promotion delivery efficiency and income, and enhancing the attractiveness and competitiveness of the promotion.

[0077] It should be noted that the above two evaluation points can be applied simultaneously in the material generation method. When the two evaluation points are applied simultaneously, the click-through rate evaluated by the second evaluation point can be the click-through rate of the first remade material, and at this time, the second remade material should be different from the first remade material and does not contain any illegal content. If the material recognition model or the click-through rate prediction model determines that the illegal content or the content affecting the click-through rate comes from the input copywriting information, the user can also be required to re-enter new copywriting information. Of course, other material evaluation points can be added as needed.

[0078] In an alternative embodiment, as Figure 7 shown, the method further includes: S700, return the target material to the target object.

[0079] S702, in response to the target object selecting the target material, save the target material and the corresponding intermediate material, copywriting information, and / or material background.

[0080] The target material can be returned to the target object in multiple formats such as JPG, PNG, etc. In addition to saving the above information, some key parameters and configuration information during the material generation process, such as the parameter settings of the model, can also be saved. In some embodiments, the usage records of the target material, such as usage time, usage scenario, delivery effect, etc., can also be saved to further optimize the material generation strategy and improve the generation quality and effect of the material. It should be noted that the location for saving the data can be in the cloud or on the local storage device of the target object.

[0081] After generating the target material, the user can save the relevant data. Thus, the material generation efficiency can be further improved, and at the same time, data support is provided for subsequent optimization and improvement, facilitating the subsequent reuse of the user for the expected solutions and better meeting the user's needs.

[0082] To make this application easier to understand, the following provides an exemplary application in conjunction with Figure 8 One is provided. Among them: S11, the promoter (i.e., the target object) uploads a SKU picture A (i.e., the initial material); S12, determine whether the material quality of picture A is lower than the preset standard; If so, go to step S13, otherwise go to step S14; S13, optimize picture A to obtain an optimized picture B (i.e., the optimized material); S14, input picture A / picture B into the material processing model to eliminate the elements irrelevant to the commodity main body O of picture A / picture B, and obtain picture C (i.e., the intermediate material); S15, the promoter selects picture C and inputs the relevant information (i.e., the copywriting information) corresponding to picture C; S16, the material generation model generates a material background adapted to picture C according to picture C and the relevant information; S17, the material generation model combines picture C into the adapted position of the material background to obtain picture D (i.e., the target material); S18, the material recognition model recognizes whether there is any illegal content in picture D; If so, go to step S16 to obtain a new picture D, otherwise go to step S19; S19, the click-through rate prediction model predicts whether the click-through rate of picture D is lower than the preset value; If so, go to step S16 to obtain a new picture D, if not, go to step S20; S20, return picture D to the promoter, the promoter selects picture D, and saves picture D, picture C, the relevant information corresponding to picture C, and the material background to the local.

[0083] Embodiment Two Figure 9 Schematically shows a block diagram of a material generation device according to Embodiment Two of the present application. The device can be divided into one or more program modules. One or more program modules are stored in a storage medium and are executed by one or more processors to complete the embodiments of the present application. The program modules referred to in the embodiments of the present application refer to a series of computer program instruction segments that can complete specific functions. The following description will specifically introduce the functions of each program module in this embodiment. As Figure 9 shown, the device 1000 may include: a first acquisition module 1100, a second acquisition module 1200, and a generation module 1300, where: The first acquisition module 1100 is used to acquire the initial material; A second acquisition module 1200, configured to obtain intermediate materials including target elements based on the initial materials through a pre-trained material processing model; A generation module 1300, configured to use the intermediate materials and the copywriting information for the intermediate materials as model inputs, and generate target materials through a pre-trained material generation model; wherein, the target materials include the target elements and the copywriting information.

[0084] As an optional embodiment, the second acquisition module is further configured to: Determine the material quality of the initial materials; In the case that the material quality of the initial materials is lower than a preset threshold, process the initial materials to obtain optimized materials; Use the optimized materials as model inputs, and generate the intermediate materials according to the optimized materials through the material processing model.

[0085] As an optional embodiment, the generation module is further configured to: Generate a material background based on the copywriting information and adapted to the intermediate materials through the material generation model; Combine the intermediate materials to the material adaptation position of the material background to obtain the target materials.

[0086] As an optional embodiment, the apparatus 1000 further includes a material recognition module, configured to: Input the target materials into a pre-trained material recognition model to obtain the recognition result of the target materials; wherein, the material recognition model is used to determine whether there is any illegal content in the target materials; In the case that there is illegal content in the target materials, use the intermediate materials and the copywriting information as model inputs, and generate first re-made materials through the material generation model, where the first re-made materials are different from the target materials.

[0087] As an optional embodiment, the apparatus 1000 further includes a click-through rate prediction module, configured to: Input the target materials into a pre-trained click-through rate prediction model to output the predicted click-through rate of the target materials through the click-through rate prediction model; In the case that the predicted click-through rate is lower than a preset value, use the intermediate materials and the copywriting information as model inputs, and generate second re-made materials through the material generation model, where the second re-made materials are different from the target materials.

[0088] As an optional embodiment, the apparatus 1000 further includes a saving module, configured to: Return the target materials to the target object; In response to the target object selecting the target material, save the target material and the corresponding intermediate material, copywriting information, and / or material background.

[0089] Embodiment III Figure 10 Schematically shows a hardware architecture diagram of a computer device 10000 suitable for implementing the material generation method according to Embodiment III of the present application. In some embodiments, the computer device 10000 may be a terminal device such as a smart phone, a wearable device, a tablet computer, a personal computer, a vehicle-mounted terminal, a game console, a virtual device, a workbench, a digital assistant, a set-top box, a robot, etc. In other embodiments, the computer device 10000 may be a rack server, a blade server, a tower server, or a cabinet server (including an independent server or a server cluster composed of multiple servers), etc. As Figure 10 shown, the computer device 10000 includes, but is not limited to: a memory 10010, a processor 10020, and a network interface 10030 that can be communicatively linked to each other through a system bus. Among them: The memory 10010 includes at least one type of computer-readable storage medium. The readable storage medium includes flash memory, a hard disk, a multimedia card, a card-type memory (such as an SD or DX memory), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 10010 may be an internal storage module of the computer device 10000, such as the hard disk or memory of the computer device 10000. In other embodiments, the memory 10010 may also be an external storage device of the computer device 10000, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc., equipped on the computer device 10000. Of course, the memory 10010 may also include both the internal storage module and the external storage device of the computer device 10000. In this embodiment, the memory 10010 is generally used to store the operating system and various application software installed on the computer device 10000, such as the program code of the material generation method. In addition, the memory 10010 may also be used to temporarily store various data that have been output or will be output.

[0090] The processor 10020 can be a Central Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or other chips in some embodiments. The processor 10020 is generally used to control the overall operation of the computer device 10000, such as performing control and processing related to data interaction or communication with the computer device 10000. In this embodiment, the processor 10020 is used to run the program code stored in the memory 10010 or process data.

[0091] The network interface 10030 may include a wireless network interface or a wired network interface. The network interface 10030 is generally used to establish a communication link between the computer device 10000 and other computer devices. For example, the network interface 10030 is used to connect the computer device 10000 to an external terminal through a network, and establish a data transmission channel and a communication link between the computer device 10000 and the external terminal. The network can be an Intranet, the Internet, Global System of Mobile communication (GSM for short), Wideband Code Division Multiple Access (WCDMA for short), 4G network, 5G network, Bluetooth, Wi-Fi and other wireless or wired networks.

[0092] It should be noted that Figure 10 Only the computer device with components 10010 - 10030 is shown, but it should be understood that it is not required to implement all the shown components, and more or fewer components can be alternatively implemented.

[0093] In this embodiment, the material generation method stored in the memory 10010 can also be divided into one or more program modules and executed by one or more processors (such as the processor 10020) to complete the embodiments of the present application.

[0094] Embodiment 4 The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the material generation method in the embodiment are implemented.

[0095] In this embodiment, the computer-readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the computer-readable storage medium may be an internal storage unit of a computer device, such as the hard disk or memory of the computer device. In other embodiments, the computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc. equipped on the computer device. Of course, the computer-readable storage medium may also include both the internal storage unit and the external storage device of the computer device. In this embodiment, the computer-readable storage medium is generally used to store the operating system installed on the computer device and various application software, such as the program code of the material generation method in the embodiment. In addition, the computer-readable storage medium can also be used to temporarily store various data that have been output or will be output.

[0096] Embodiment 5 The embodiment of the present application also provides a computer program product, including a computer program, which when executed by a processor implements the method in the above embodiment.

[0097] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the embodiments of the present application can be implemented by a general computer device. They can be concentrated on a single computer device or distributed on a network composed of multiple computer devices. Optionally, they can be implemented by program codes executable by the computer device, so that they can be stored in a storage device and executed by the computer device. And in some cases, the steps shown or described can be executed in a different order than here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module to implement. Thus, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0098] It should be noted that the above are only the preferred embodiments of the present application, and do not limit the patent protection scope of the present application. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. A material generation method, characterized in that: The method comprises: Get initial materials; Acquire intermediate material including the target element based on the initial material by using a pre-trained material processing model; The intermediate material and the copy information for the intermediate material are used as model inputs, and a target material is generated through a pre-trained material generation model; wherein the target material includes the target element and the copy information.

2. The method according to claim 1, characterized in that Acquiring an intermediate material including a target element based on the initial material includes: Determining the material quality of the initial material; When the material quality of the initial material is lower than a preset standard, processing the initial material to obtain an optimized material; The optimized material is used as a model input, and the intermediate material is generated according to the optimized material through the material processing model.

3. The method according to claim 1, characterized in that The intermediate material and the copywriting information for the intermediate material are used as model inputs, and the target material is generated by a pre-trained material generation model, including: Generate a material background based on the copy information and adapted to the intermediate material through the material generation model; The intermediate material is combined into a material adaptation position of the material background to obtain the target material.

4. The method according to claim 1, characterized in that: The method further comprises: Inputting the target material into a pre-trained material recognition model to obtain a recognition result of the target material; wherein the material recognition model is used to determine whether the target material contains illegal content; In the case that the target material contains illegal content, the intermediate material and the text information are used as model inputs, and a first remade material is generated through the material generation model, where the first remade material is different from the target material.

5. The method according to claim 1, characterized in that The method further comprises: Inputting the target material into a pre-trained click rate prediction model to output a predicted click rate of the target material through the click rate prediction model; When the predicted click rate is lower than a preset value, the intermediate material and the text information are used as model inputs, and a second remade material is generated through the material generation model, and the second remade material is different from the target material.

6. The method according to claim 1, characterized in that The method further comprises: Returning the target material to the target object; In response to the target object selecting the target material, the target material and corresponding intermediate materials, text information and / or material background are saved.

7. A material generation device, characterized in that: The device comprises: A first acquisition module, used to acquire initial materials; A second acquisition module, configured to acquire intermediate material including a target element based on the initial material by using a pre-trained material processing model; A generation module is used to take the intermediate material and the copy information for the intermediate material as model input, and generate a target material through a pre-trained material generation model; wherein the target material includes the target element and the copy information.

8. A computer device, characterized in that: include: at least one processor; and a memory communicatively coupled to the at least one processor; wherein: The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images