A makeup scheme recommendation method, device, equipment and storage medium

Abstract

Embodiments of the present application disclose a makeup scheme recommendation method and device, equipment and a storage medium, related embodiments can be applied to cloud technology, artificial intelligence, intelligent transportation and various scenes, to improve the accuracy of makeup scheme recommendation. The method of the embodiments of the present application comprises: in response to a first instruction triggered for a current video playing interface, obtaining an original face image of a target object, obtaining a target character face image matched with the original face image, wherein the target character face image is derived from a character of a first video played by the current video playing interface, matching a plurality of character makeup schemes corresponding to the original face image and the target character face image, obtaining a makeup matching degree, determining a target character makeup scheme from the plurality of character makeup schemes according to the makeup matching degree, and pushing the target character makeup scheme.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a method, apparatus, device, and storage medium for recommending makeup schemes. Background Technology

[0002] With image becoming increasingly important and the rise of the "appearance economy," people's demand for beauty and skincare products is growing stronger. With a wide variety of beauty products and countless makeup tutorials available, it is becoming increasingly difficult for beauty consumers to make a choice.

[0003] Currently, makeup tutorials and videos are readily available on online platforms, allowing people to follow the steps to complete their makeup routine. However, due to the vast array of cosmetics on the market, the products available to users often differ from those included in the tutorials. Even if the colors and usage are the same, the desired results may not be achieved. Furthermore, many beginners lack professional knowledge about cosmetics and makeup techniques, leaving them unsure where to begin when faced with generic skincare advice, a dazzling array of beauty products, and overly similar makeup tutorials. They urgently need a makeup style tailored to their individual needs. Summary of the Invention

[0004] This application provides a method, apparatus, device, and storage medium for recommending makeup schemes. It is used to digitally express the degree of fit between the target object and each character's makeup scheme by using the makeup matching degree between the original face image and the target character's face image corresponding to several character makeup schemes. This allows for a more accurate acquisition of a target character makeup scheme with film and television characteristics that is suitable for the target object.

[0005] This application provides a method for recommending makeup schemes, including:

[0006] In response to the first command triggered by the current video playback interface, the original face image of the target object is obtained;

[0007] Obtain the target character's face image that matches the original face image, wherein the target character's face image is derived from the character in the first video played on the current video playback interface;

[0008] The original face image is matched with several character makeup schemes corresponding to the target character's face image to obtain the makeup matching degree;

[0009] Based on the makeup matching degree, the target character's makeup scheme is determined from several character makeup schemes, and the target character's makeup scheme is pushed to the user.

[0010] This application also provides a device for recommending makeup schemes, comprising:

[0011] The acquisition unit is used to acquire the original face image of the target object in response to a first instruction triggered for the current video playback interface;

[0012] The acquisition unit is also used to acquire a target character face image that matches the original face image, wherein the target character face image is derived from the character in the first video played on the current video playback interface;

[0013] The processing unit is used to match the original face image with several character makeup schemes corresponding to the target character face image to obtain the makeup matching degree;

[0014] The determination unit is used to identify the target character's makeup scheme from several character makeup schemes based on the makeup matching degree, and then push the target character's makeup scheme.

[0015] In one possible design, in another implementation of the embodiments of this application,

[0016] The acquisition unit is also used to acquire the basic object face image, the character makeup sample image, and the basic attribute features of the character corresponding to the character in the character makeup sample image. The character makeup sample image is any frame face image of each film and television character extracted from the film and television works, and the character makeup sample image has a corresponding makeup label.

[0017] The processing unit is also used to extract the facial contour features of the base object from the base object's facial image, and to extract the character's makeup features and the character's facial contour features from the makeup sample image.

[0018] The processing unit is also used to input the basic object's facial contour features, the character's basic attribute features, the character's makeup features, and the character's facial contour features into the makeup prediction model, and output the makeup prediction probability through the makeup prediction model.

[0019] The processing unit is also used to update the model parameters of the makeup prediction model based on the makeup prediction probability and the makeup label;

[0020] The processing unit can be used to: input the original face image into the makeup prediction model, output the makeup prediction probability corresponding to each character's makeup scheme through the makeup prediction model, and determine the makeup matching degree between the original face image and each character's makeup scheme based on the makeup prediction probability.

[0021] In one possible design, in another implementation of the embodiments of this application, the target character's makeup scheme includes makeup tutorial information and makeup product information;

[0022] Makeup tutorial information includes several makeup steps, and each makeup step includes at least one of the following: instructions for using the makeup products and a makeup effect image.

[0023] Makeup product information includes base makeup product information and color makeup product information, which appears in makeup tutorial information.

[0024] In one possible design, in another implementation of the embodiments of this application, the acquisition unit may specifically be used for:

[0025] Perform a facial scan on the target object to obtain the original face image; or

[0026] The system identifies the face photos uploaded by the target object to obtain the original face image.

[0027] In one possible design, in another implementation of the embodiments of this application, the acquisition unit may specifically be used for:

[0028] Extract facial feature information from the original face image, and extract skin state information from the original face image;

[0029] Based on facial feature information and skin condition information, calculate the facial matching degree between the original face image and the face images of each character in the first video played on the current video playback interface.

[0030] Based on facial matching accuracy, the target character's facial image is determined from the facial images of each character in the first video.

[0031] In one possible design, in another implementation of the embodiments of this application, the acquisition unit may specifically be used for:

[0032] Facial feature extraction is performed on the original face image to obtain the original facial feature map;

[0033] Facial feature points are extracted from the original face image, and the extracted facial feature points are connected to obtain a facial contour feature map;

[0034] The original facial feature map and the facial contour feature map are superimposed to obtain the target facial feature map.

[0035] The facial features of the target face are decomposed to obtain facial feature information.

[0036] In one possible design, in another implementation of the embodiments of this application, the acquisition unit may specifically be used for:

[0037] The original face image is subjected to intrinsic image decomposition processing to obtain the specular intrinsic layer, diffuse intrinsic layer and skin color intrinsic layer;

[0038] The oiliness of facial skin is determined by the ratio of the highlight intrinsic layer to the diffuse intrinsic layer.

[0039] The roughness is determined based on the gradient value of the diffuse intrinsic layer;

[0040] Pigment concentration was identified in the intrinsic layer of skin color to obtain hemoglobin concentration and melanin concentration.

[0041] In one possible design, in another implementation of the embodiments of this application,

[0042] The acquisition unit is also used to read the database and retrieve the skin quality report generation template from the database;

[0043] The acquisition unit is also used to acquire the first skin care product category corresponding to the oiliness of the human face skin according to the preset correspondence between oiliness and skin care product category, and to acquire the second skin care product category corresponding to roughness according to the preset correspondence between roughness and skin care product data, and to acquire the third skin care product category corresponding to pigment concentration according to the preset correspondence between hemoglobin concentration and melanin concentration and skin care product data.

[0044] The processing unit is also used to take the intersection of the first skin care category, the second skin care category, and the third skin care category as the target skin care product set;

[0045] The processing unit is also used to generate a skin condition report corresponding to the target object based on the skin condition information according to the report generation template;

[0046] The processing unit is also used to push skin condition reports and target skincare product sets to the target object.

[0047] In one possible design, in another implementation of the embodiments of this application, the determining unit may specifically be used for:

[0048] The facial matching degree between the target character's face image and the original face image is weighted and summed with the makeup matching degree corresponding to each character's makeup scheme to obtain the makeup score corresponding to each character's makeup scheme.

[0049] Based on the makeup score, several character makeup schemes are filtered to obtain the target character's makeup scheme and push it to the user.

[0050] In one possible design, in another implementation of the embodiments of this application,

[0051] The acquisition unit is also used to scan the face of the target object and acquire the current face scan image;

[0052] The acquisition unit is also used to extract the currently made-up area from the current face scan image based on the unmade-up face image of the target object;

[0053] The processing unit is also used to compare the currently applied makeup area with the standard makeup area of ​​the face image in the target character's makeup scheme to obtain the comparison result;

[0054] The processing unit is also used to send a prompt to the target object to proceed to the next stage of makeup if the comparison results are consistent.

[0055] The processing unit is also used to send a prompt to the target object that the currently applied makeup area is inconsistent with the target character's makeup scheme if the comparison result is inconsistent.

[0056] Another aspect of this application provides a computer device, including: a memory, a transceiver, a processor, and a bus system;

[0057] The memory is used to store programs;

[0058] The processor implements the methods described above when executing a program in memory;

[0059] Bus systems are used to connect memory and processor to enable communication between them.

[0060] Another aspect of this application provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the methods described above.

[0061] As can be seen from the above technical solutions, the embodiments of this application have the following advantages:

[0062] By responding to a first command triggered by the current video playback interface, the original facial image of the target object is obtained. Then, the facial image of the target character from the first video played on the current video playback interface, matching the original facial image, is acquired. Next, several character makeup schemes corresponding to the original facial image and the target character's facial image are matched to obtain a makeup matching degree. Based on this matching degree, the target character's makeup scheme is determined from these schemes and recommended. Through this method, based on a rich set of film and television character makeup schemes, the degree of compatibility between the target object and each character's makeup scheme is digitally expressed through the makeup matching degree between the original facial image and the target character's facial image. This allows for a more accurate acquisition of a film and television-specific target character makeup scheme that suits the target object and meets the target object's diverse makeup needs for film and television characters. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the architecture of the makeup scheme control system in the embodiments of this application;

[0064] Figure 2 This is a flowchart of an embodiment of the method for recommending makeup schemes in this application;

[0065] Figure 3 This is a flowchart of another embodiment of the method for recommending makeup schemes in this application;

[0066] Figure 4 This is a flowchart of another embodiment of the method for recommending makeup schemes in this application;

[0067] Figure 5 This is a flowchart of another embodiment of the method for recommending makeup schemes in this application;

[0068] Figure 6 This is a flowchart of another embodiment of the method for recommending makeup schemes in this application;

[0069] Figure 7 This is a flowchart of another embodiment of the method for recommending makeup schemes in this application;

[0070] Figure 8 This is a flowchart of another embodiment of the method for recommending makeup schemes in this application;

[0071] Figure 9 This is a flowchart of another embodiment of the method for recommending makeup schemes in this application;

[0072] Figure 10 This is a flowchart of another embodiment of the method for recommending makeup schemes in this application;

[0073] Figure 11 This is a flowchart illustrating a principle embodiment of the makeup scheme recommendation method in this application.

[0074] Figure 12 This is another schematic diagram illustrating the principle and flow of the makeup scheme recommendation method in the embodiments of this application;

[0075] Figure 13 This is another schematic diagram illustrating the principle and flow of the makeup scheme recommendation method in the embodiments of this application;

[0076] Figure 14 This is a schematic diagram of another makeup prediction model training process for the makeup scheme recommendation method in the embodiments of this application;

[0077] Figure 15This is a schematic diagram of one embodiment of the makeup scheme recommendation device in this application;

[0078] Figure 16 This is a schematic diagram of one embodiment of the computer device described in this application. Detailed Implementation

[0079] This application provides a method, apparatus, device, and storage medium for recommending makeup schemes. It is used to digitally express the degree of fit between the target object and each character's makeup scheme by using the makeup matching degree between the original face image and the target character's face image corresponding to several character makeup schemes. This allows for a more accurate acquisition of a target character makeup scheme with film and television characteristics that is suitable for the target object.

[0080] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and drawings of this application are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “corresponding to,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0081] With the rapid development of information technology, cloud technology is gradually permeating all aspects of people's lives. Cloud technology is a general term encompassing network technology, information technology, integration technology, management platform technology, and application technology based on the cloud computing business model. It can form resource pools, providing flexible and convenient on-demand access. Cloud computing technology will become a crucial support. Backend services of technical network systems require substantial computing and storage resources, such as video websites, image websites, and many portal websites. With the rapid development and application of the internet industry, every item may have its own identification mark in the future, requiring data to be transmitted to backend systems for logical processing. Data at different levels will be processed separately, and various industry data will require robust system support, which can only be achieved through cloud computing.

[0082] Cloud security refers to the collective term for security software, hardware, users, organizations, and security cloud platforms based on cloud computing business models. Cloud security integrates emerging technologies and concepts such as parallel processing, grid computing, and unknown virus behavior detection. It uses a large network of clients to monitor abnormal software behavior on the network, obtain the latest information on Trojans and malware on the internet, and send it to the server for automatic analysis and processing. Finally, solutions for viruses and Trojans are distributed to each client. The makeup recommendation method provided in this application embodiment can be implemented using cloud computing and cloud security technologies.

[0083] It should be understood that the makeup recommendation method provided in this application can be applied to fields such as cloud technology, artificial intelligence, and intelligent transportation, for scenarios such as providing makeup references or tutorials to target objects through recommended makeup schemes. For example, a makeup scheme suitable for target object A can be obtained and recommended to target object A as a daily makeup reference. As another example, a themed makeup scheme suitable for target object B can be obtained to guide target object B in completing a themed makeup look. As yet another example, a makeup scheme suitable for target object C can be obtained to assist target object C in choosing more suitable makeup products.

[0084] To address the aforementioned issues, this application proposes a method for recommending makeup schemes, which is applied to... Figure 1 Please refer to the makeup scheme control system shown. Figure 1 , Figure 1 This is a schematic diagram of the architecture of the makeup solution control system in an embodiment of this application, as shown below. Figure 1 As shown, the server, in response to a first command triggered by the current video playback interface on the terminal device, obtains the original facial image of the target object. It can then obtain the target character's facial image from the first video played on the current video playback interface, matching the original facial image. Next, it matches the original facial image with several corresponding character makeup schemes to obtain a makeup matching degree. Based on this matching degree, it determines the target character's makeup scheme from among the several schemes and pushes that scheme. Through this method, based on a rich set of film and television character makeup schemes, the degree of compatibility between the target object and each character makeup scheme can be digitally expressed through the makeup matching degree between the original facial image and the target character's facial image. This allows for a more accurate acquisition of a film and television-specific target character makeup scheme that suits the target object and meets the target object's diverse makeup needs for film and television characters.

[0085] Understandable Figure 1Only one type of terminal device is shown in the diagram. In real-world scenarios, many more types of terminal devices can participate in the data processing. These include, but are not limited to, mobile phones, computers, smart voice interaction devices, smart home appliances, and in-vehicle terminals. The specific number and types depend on the actual scenario and are not limited here. Furthermore, Figure 1 The diagram shows one server, but in real-world scenarios, multiple servers can be involved, especially in scenarios involving multi-model training and interaction. The number of servers depends on the specific scenario and is not limited here.

[0086] It should be noted that in this embodiment, the server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. Terminal devices and servers can be directly or indirectly connected via wired or wireless communication, and terminal devices and servers can be connected to form a blockchain network; this application does not impose any limitations on this.

[0087] To address the aforementioned issues, this application proposes a method for recommending makeup schemes. This method is generally executed by a server or terminal device, and correspondingly, the device for recommending makeup schemes is generally located in the server or terminal device.

[0088] It is understood that, as disclosed in this application, the recommended methods, apparatus, devices, and storage media for makeup solutions can include multiple servers or terminal devices forming a blockchain, with each server or terminal device acting as a node on the blockchain. In practical applications, data sharing between nodes within the blockchain is required, and each node can store facial image data, makeup data, etc.

[0089] The following section will introduce the recommended methods for the makeup scheme in this application. Please refer to [link / reference needed]. Figure 2 One embodiment of the makeup scheme recommendation method in this application includes:

[0090] In step S101, in response to the first instruction triggered for the current video playback interface, the original face image of the target object is obtained;

[0091] In this embodiment, as Figure 12As shown, when there is a character makeup that the target object is interested in in the current video playback interface, the target object can trigger a character makeup matching instruction, i.e., a first instruction, through the current video playback interface on the terminal device, so that the server or terminal device can respond to the first instruction triggered for the current video playback interface and obtain the original face image of the target object according to the first instruction.

[0092] Specifically, the target object can be a user browsing or viewing the current video playback interface using a terminal device. The original facial image can be a facial image obtained through scanning devices such as cameras, a pre-uploaded photo of the target object without makeup, or a facial image; no specific restrictions are imposed here.

[0093] Specifically, such as Figure 11 As shown, when a target user has a desired makeup look for a character in the current video playback interface, the target user can trigger a makeup matching command (the first command) through the current video playback interface on their terminal device. This allows the terminal device to respond to the first command triggered by the current video playback interface. Considering that facial recognition involves user privacy and network information security, this embodiment will remind the target user whether they agree to the data collection via pop-ups or timely messages to optimize their experience. When the target user agrees, they can select the facial collection operation through the facial scanning interface of the client installed on the terminal device according to the first command. The facial image of the target user will be collected and sent to the server, allowing the server to receive the target user's original facial image. It is understood that the target user can also perform personal information selection or entry operations through the information authorization interface or entry interface provided by the client installed on the terminal device. This generates corresponding basic object information and makeup preference information, which is then sent to the makeup scheme recommendation server, allowing the server to obtain the target user's basic object information and makeup preference information.

[0094] Alternatively, the target can send a makeup recommendation instruction carrying the target's identifier, i.e., the first instruction, to the server through a client installed on the terminal device. This allows the server to respond to the first instruction and, based on the first instruction, retrieve the original face image, basic object information, and makeup preference information corresponding to the target's identifier from the database. Alternatively, the original face image, basic object information, and makeup preference information can be obtained through other methods, without specific restrictions here.

[0095] The basic target information can specifically include the target's age, makeup scenario (such as the occasion), role preference, product brand preference, and consumption preference (such as purchasing power), as well as other target information, without specific limitations. Makeup preference information can specifically include colors such as foundation, blush, and eyeshadow, themes such as date makeup, holiday makeup, and light makeup for interviews or commuting, or shapes such as eyebrow shape, lip shape, or nose shape, and can also include other makeup preferences, without specific limitations.

[0096] In step S102, a target character face image matching the original face image is obtained, wherein the target character face image is derived from the character in the first video played on the current video playback interface;

[0097] In this embodiment, as Figure 12 As shown, after obtaining the original face image, the target character's face image that matches the original face image can be matched from the characters in the first video played on the current video playback interface. This allows for the intelligent matching of film and television makeup to the target character based on the target character's face image that has a high degree of fit with the original face image, thereby improving the accuracy of the target character's makeup scheme recommendation to a certain extent.

[0098] The target character's face image is derived from the character in the first video played on the current video playback interface. It can be understood as a character's face image that is similar to the face shape or facial features in the original face image of the target object.

[0099] Specifically, since facial features and the distribution of facial features are the main factors in recognizing faces, in order to obtain the target character's face image that best matches the facial features and the distribution of facial features in the original face image, after obtaining the original face image, the target character's face image that matches the original face image is obtained. Specifically, this can be done by using an object recognition model or a face recognition model to extract the facial feature information of the target object from the obtained original face image. Based on the extracted facial feature information, the image similarity or facial matching degree between the original face image and the face images of each character in the first video played on the current video playback interface can be calculated. Specifically, this can be done using cosine similarity calculation formulas or Euclidean distance calculation formulas, etc., without specific limitations here. Then, the face image of the character with the highest facial matching degree can be determined as the target character's face image. Alternatively, other methods can be used to obtain the target character's face image that matches the original face image, such as using facial feature information and skin state information to more accurately match the target character's face image, without specific limitations here.

[0100] In step S103, the original face image is matched with several character makeup schemes corresponding to the target character face image to obtain the makeup matching degree.

[0101] In this embodiment, as Figure 12 As shown, after obtaining the target character's face image, the original face image can be matched with several character makeup schemes corresponding to the target character's face image to obtain the makeup matching degree. This allows the subsequent selection of several character makeup schemes based on the obtained makeup matching degree, thereby accurately obtaining the target character makeup scheme with a high degree of suitability to the target object, which can improve the accuracy of recommending the target character makeup scheme to a certain extent.

[0102] Several of the character makeup schemes are derived from all the film and television works in which the actor playing the target character's facial image has portrayed the character. Due to the requirements of different scenes or situations in the film and television works, each character played by the actor can correspond to one or more character makeup schemes. Each character makeup scheme is extracted by analyzing a frame of the character's facial image from a film or television work.

[0103] Specifically, after obtaining the target character's face image, the original face image can be matched with several character makeup schemes corresponding to the target character's face image. Specifically, the original face image can be input into a makeup prediction model, and the makeup prediction model can output the makeup prediction probability corresponding to each character makeup scheme. The makeup prediction probability is then used as the makeup matching degree between the original face image and the character makeup scheme.

[0104] In step S104, based on the makeup matching degree, the target character makeup scheme is determined from several character makeup schemes and the target character makeup scheme is pushed.

[0105] In this embodiment, as Figure 12 As shown, after obtaining the makeup matching degree between the original face image and each character's makeup scheme, the most suitable target character makeup scheme can be selected from several character makeup schemes based on the makeup matching degree, and the target character makeup scheme can be pushed. This can more accurately help the target to obtain a target character makeup scheme with film and television characteristics that is suitable for the target, and can also meet the target's makeup needs for diverse film and television characters.

[0106] Specifically, after obtaining the makeup matching degree between the original face image and each character's makeup scheme, since a higher makeup matching degree can be understood as a higher degree of compatibility between the target object and the character's makeup scheme, the makeup matching degrees can be compared pairwise, and the character's makeup scheme with the highest makeup matching degree can be taken as the target character's makeup scheme.

[0107] Furthermore, in order to better meet the makeup needs of the target audience for diverse film and television roles, the matching degree or hit rate between the obtained basic object information and makeup preference information and the makeup tags of each character's makeup scheme can be calculated based on the basic object information and makeup preference information. Then, the makeup prediction probability and hit rate can be weighted according to preset weights to obtain several weighted sum values. Since the larger the weighted sum value, it can be understood that the higher the fit between the target object and the character's makeup scheme, the character's makeup scheme corresponding to the largest weighted sum value can be taken as the target character's makeup scheme. Other methods can also be used to determine the target character's makeup scheme, and no specific restrictions are imposed here.

[0108] The makeup tags for each character's makeup scheme are used to indicate the makeup theme, such as a natural makeup look, peach blossom makeup, maple leaf makeup, maid makeup, and smoky makeup, etc. Other themes can also be represented, without specific restrictions here.

[0109] Specifically, the matching degree between the obtained basic object information and makeup preference information and the makeup tags of each character's makeup scheme can be achieved by using semantic recognition or topic recognition models to perform semantic recognition or topic recognition on the basic object information and makeup preference information. The corresponding semantic recognition results or topic recognition results can be obtained. Then, the semantic similarity between the semantic recognition results or topic recognition results and the makeup tags of each makeup scheme can be calculated. The obtained semantic similarity can then be used as the matching degree.

[0110] In this application embodiment, a method for recommending makeup schemes is provided. Through the above method, based on a wealth of film and television character makeup schemes, the degree of compatibility between the target object and each character makeup scheme is digitally expressed by the makeup matching degree between the original face image and the target character face image corresponding to several character makeup schemes. This allows for more accurate acquisition of target character makeup schemes with film and television characteristics that are compatible with the target object, and also meets the target object's makeup needs for diverse film and television characters.

[0111] Optionally, in the above Figure 2 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application, such as... Figure 3 As shown, before step S103 matches the original face image with several character makeup schemes corresponding to the target character face image to obtain the makeup matching degree, the method further includes: steps S301 to S304, and step S103 includes: step S305.

[0112] In step S301, the basic object face image, the character makeup sample image, and the basic attribute features of the character corresponding to the character in the character makeup sample image are obtained. The character makeup sample image is any frame face image of each film and television character extracted from the film and television works, and the character makeup sample image has a corresponding makeup label.

[0113] In step S302, the basic object face contour features are extracted from the basic object face image, and the character makeup features and character face contour features are extracted from the makeup sample image.

[0114] In step S303, the basic object's facial contour features, the character's basic attribute features, the character's makeup features, and the character's facial contour features are input into the makeup prediction model, and the makeup prediction model outputs the makeup prediction probability.

[0115] In step S304, the model parameters of the makeup prediction model are updated using the makeup prediction probability and makeup label;

[0116] In step S305, the original face image is input into the makeup prediction model, the makeup prediction model outputs the makeup prediction probability corresponding to each character's makeup scheme, and the makeup matching degree between the original face image and each character's makeup scheme is determined based on the makeup prediction probability.

[0117] In this embodiment, as Figure 12As shown, before matching the original face image with several character makeup schemes corresponding to the target character's face image to obtain the makeup matching degree, a makeup scheme recommendation database can be read. From this database, the base object face image, character makeup sample images, and the basic attribute features of the characters corresponding to the characters in the character makeup sample images can be obtained. Then, base object face contour features can be extracted from the base object face image, and character makeup features and character face contour features can be extracted from the makeup sample images. These base object face contour features, character basic attribute features, character makeup features, and character face contour features can be input into a makeup prediction model. The makeup prediction model outputs a makeup prediction probability, which can then be used to predict the makeup match. The predicted probabilities and makeup labels update the model parameters of the makeup prediction model. The original face image is input into the makeup prediction model, which outputs the makeup prediction probability corresponding to each character's makeup scheme. Based on the makeup prediction probability, the makeup matching degree between the original face image and each character's makeup scheme is determined. The model can be trained with high prediction accuracy by using the basic object's facial contour features, the character's basic attribute features, the character's makeup features, and the character's facial contour features. This allows the trained makeup prediction model to quickly and accurately obtain the makeup matching degree between the target object and each character's makeup scheme, thereby improving the efficiency and accuracy of recommended makeup schemes to a certain extent.

[0118] The character makeup sample images are any frame of facial image captured from each film and television character, and each image is tagged with a makeup label. The basic attribute features corresponding to the characters in the makeup sample images are obtained by accessing existing object information databases, such as a celebrity database, to obtain basic attributes like the age or gender of the actor playing the role. This allows for further matching with the target object's age or gender. It's understandable that since film and television characters may have significant differences from the actor's age or gender (e.g., a man dressed as a woman, a woman dressed as a man, or a young girl dressed as an old person), these differences are filtered during the acquisition of the makeup sample images to ensure that the obtained makeup features match the actor's current age and gender. The makeup features refer to the eye makeup, lip makeup, and eyebrow makeup in any frame of facial image captured for each film and television character. The facial contour features refer to the facial contours and features extracted from the captured makeup sample images after weakening the makeup effects, aiming to restore the character's natural appearance as much as possible.

[0119] Specifically, such as Figure 14As shown, by reading the makeup scheme recommendation database, basic object face images and character makeup sample images can be obtained from the database. Furthermore, the basic attributes of the character corresponding to the character makeup sample image, such as age, gender, eye shape, or face shape, can be searched from the object information database through the character information call interface. In other words, the basic attribute features of the character corresponding to the character in the character makeup sample image are obtained.

[0120] Furthermore, after obtaining the basic object face image and the character makeup sample image, the basic object face contour features can be extracted from the basic object face image. The facial feature information extraction module can be used to extract basic facial features from the obtained basic object face image to obtain a basic facial feature map. Specifically, the basic face image can be processed by several fully connected layers or pooling layers in a convolutional neural network (CNN) to obtain the basic facial feature map. Alternatively, other deep neural network frameworks such as SSD convolutional neural networks can be used to extract basic facial features from the basic object face image. No specific restrictions are imposed here. Then, the ASM algorithm can be used to detect anchor points of facial contours, eyebrows, eyes, nose, or mouth in the face of the basic object face image as basic facial feature points. Then, the detected anchor points can be connected to form linear contours, such as face contours and facial feature contours, to display the area of ​​facial features and obtain the basic facial contour feature map.

[0121] Furthermore, after obtaining the basic facial feature map and the basic facial contour feature map, the basic facial contour feature map containing linear contour anchor points can be aligned and superimposed with the basic facial feature map containing the original facial features through operations such as scaling, transformation, and translation. This can optimize pixels, reduce the influence of cluttered pixels, and improve image quality, resulting in a high-quality, information-rich basic object facial feature map. Then, the obtained basic object facial feature map can be processed by facial feature decomposition to obtain detailed and accurate facial feature information, such as features like almond-shaped eyes, double eyelids, high and low nose bridges, thick and thin lips, and thick and thin eyebrows, which are the basic object facial contour features.

[0122] Furthermore, character makeup features and character facial contour features are extracted from the character makeup sample images. Specifically, the character makeup sample images can be processed by makeup effect processing models and methods to extract makeup and weaken makeup effects, resulting in character makeup features and weakened sample images that restore the character's natural makeup as much as possible. Then, character facial contour features can be extracted from the weakened sample images, which can be done in a manner similar to extracting basic object facial contour features from basic object facial images, and will not be elaborated further here. This enables better matching of similar character facial contour features to the target object in subsequent steps.

[0123] Furthermore, AI deep learning and ensemble learning are performed on the basic object's facial contour features, the character's basic attribute features, the character's makeup features, and the character's facial contour features. Specifically, the basic object's facial contour features, the character's basic attribute features, the character's makeup features, and the character's facial contour features are input into the makeup prediction model. The makeup prediction model outputs the makeup prediction probability. Then, the model parameters of the makeup prediction model can be updated using the makeup prediction probability and the makeup label until the model parameters converge, thus obtaining a trained makeup prediction model.

[0124] Furthermore, after obtaining the trained makeup prediction model, the original face image can be input into the makeup prediction model. The makeup prediction model outputs the makeup prediction probability corresponding to each character's makeup scheme. It can be understood that the higher the makeup prediction probability, the higher the makeup matching degree between the target object and the character's makeup scheme. Therefore, the value of the makeup prediction probability can be used as the makeup matching degree between the target object and each character's makeup scheme.

[0125] Optionally, in the above Figure 2 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application,

[0126] The target character's makeup plan includes makeup tutorials and makeup product information;

[0127] Makeup tutorial information includes several makeup steps, and each makeup step includes at least one of the following: instructions for using the makeup products and a makeup effect image.

[0128] Makeup product information includes base makeup product information and color makeup product information, which appears in makeup tutorial information.

[0129] In this embodiment, the makeup tutorial information includes several makeup application steps. Each step includes at least one of the following: instructions for using the makeup products and a makeup effect image. The makeup effect image can be a before-and-after comparison of each makeup product, or it can be an image showing the makeup effect after applying the makeup products, combined with the product instructions. The makeup effect image can be a static image, such as a poster-style graphic or notes, or a dynamic image, such as an animation or PowerPoint presentation breaking down the makeup application steps; no specific limitations are imposed here. The makeup product information includes base makeup product information such as primer, makeup base, sunscreen, and foundation, as well as color makeup product information such as eyeshadow, eyeliner, eyebrow pencil, or lipstick. It is understood that not all makeup product information will appear in the makeup tutorial information; only makeup product information related to the target character's makeup in the makeup scheme or associated with the target character can appear in the makeup tutorial information.

[0130] Specifically, after obtaining the makeup plan for the target character, the makeup tutorial information in the makeup plan can be displayed on the display interface of the terminal device used by the target audience through decomposed animations. The makeup step information can guide the target audience to use makeup products to apply makeup in terms of actions, positions, angles, etc. For example, when applying foundation and concealer, the technique of dabbing, smearing, or brushing can be shown depending on whether the concealer is for acne marks, dark circles, or eye bags. When shaping eyebrows, the position of the brow head, brow peak, and brow tail can be determined, or the stray hairs that need to be removed can be highlighted with color. When applying highlighter and contouring, the local area can be outlined with dotted circles, etc.

[0131] Optionally, in the above Figure 2 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application, such as... Figure 4 As shown, step S101, obtaining the original face image of the target object, includes:

[0132] In step S401, a facial scan is performed on the target object to obtain an original face image; or

[0133] In step S402, the face photo uploaded by the target object is identified to obtain the original face image.

[0134] In this embodiment, when a target object in the current video playback interface has a character makeup that interests them, the target object can trigger a first instruction through the current video playback interface on the terminal device, and perform a facial scan on the target object according to the first instruction to obtain the original face image, or recognize the face photo uploaded by the target object to obtain the original face image. This allows for a better acquisition of a character makeup scheme that matches the original face image, thereby improving the accuracy of the character makeup scheme recommendation to a certain extent.

[0135] Specifically, when a target object has a character's makeup that interests them in the current video playback interface, the target object can trigger a first command through the current video playback interface on the terminal device. At the same time, considering that facial recognition involves user privacy and network information security, this embodiment will remind the target object whether to agree to the collection in the form of pop-up windows or timely information to optimize the target object's experience. Then, after the target object agrees, the terminal device can start a scanning device or a shooting device to scan or recognize the target object's face according to the first command to obtain the target object's original facial image, and send the collected original facial image to the server so that the server can receive and store the target object's original facial image.

[0136] Alternatively, a terminal device can receive a facial photo, such as a bare-faced photo, uploaded by the target object, and send the bare-faced photo uploaded by the target object to the server. The server can then use an object recognition model or a facial recognition model to identify the bare-faced photo of the target object and obtain an original facial image containing information such as the target object's face shape, eye shape, nose shape, lip shape, eyebrow shape, and the three courts and five eyes.

[0137] Optionally, in the above Figure 2 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application, such as... Figure 5 As shown, step S102, obtaining the target character's face image that matches the original face image, includes:

[0138] In step S501, facial feature information is extracted from the original face image, and skin state information is extracted from the original face image;

[0139] In step S502, based on facial feature information and skin condition information, the facial matching degree between the original face image and the face images of each character in the first video played on the current video playback interface is calculated.

[0140] In step S503, the target character's face image is determined from the face images of each character in the first video based on the face matching degree.

[0141] In this embodiment, as Figure 12 As shown, since facial features and the distribution of facial features are the main factors in makeup effects, in order to obtain a makeup scheme that matches the facial features and distribution of the target object, after obtaining the original face image of the target object, a facial feature extraction module can be used to extract facial features from the original face image to obtain the target object's facial feature information. Simultaneously, since skin condition is a key factor in makeup presentation, in order to obtain a makeup scheme that matches the target object's skin type, a skin condition extraction module can be used to extract skin features from the original face image to obtain the target object's skin condition information. Then, in the makeup scheme recommendation module, the similarity between the target object's facial feature information and skin condition information and the facial feature information of each character in the face images of the first video can be calculated as the facial matching degree between the original face image and each character's face image. Thus, the facial matching degree can be used to accurately obtain the target character's face image that best matches the target object's original face image.

[0142] Specifically, the facial feature information of the target object can be represented by the face shape, eye shape, nose shape, lip shape, eyebrow shape, and the three sections and five features of the face, or other facial features; no specific restrictions are imposed here. Skin condition information can be represented by skin color, skin texture, and skin condition, or other skin information; no specific restrictions are imposed here.

[0143] Specifically, such as Figure 11 As shown, after obtaining the original face image of the target object, facial feature information is extracted from the original face image. Specifically, the facial contour features of the face can be extracted first through the face anchoring algorithm, and then the facial features in the original face image can be decomposed through feature decomposition technology to obtain the facial features of the target object, thereby obtaining facial feature information containing facial contour features and facial features.

[0144] Furthermore, skin state information can be extracted from the original face image. Specifically, this can be achieved by using an image decomposition algorithm to decompose the original face image into a reflectance eigenmap and a luminance eigenmap. Then, illumination calibration processing can be performed on the reflectance eigenmap and luminance eigenmap respectively to obtain illuminated reflectance eigenmap and luminance eigenmap. This can effectively handle color saturation caused by overexposure, and thus reflect the true skin state of the target object as much as possible through the illuminated reflectance eigenmap and luminance eigenmap, thereby obtaining the skin state information of the target object.

[0145] Furthermore, after obtaining the facial feature information and skin state information of the target object, the similarity between the original face image and the facial feature information of each character in the face image of the first video is calculated. Specifically, this can be done by first obtaining the facial feature information of each character in the face image of the first video. The method of obtaining the facial feature information is similar to the method of extracting the facial contour features of the base object from the face image of the base object in step S302, and will not be repeated here. Then, the facial feature information of each character in the face image of the first video, the facial feature information of the target object, and the skin state information are used to calculate the similarity between the facial feature information and skin state information of the target object and the facial feature information of each character in the face image of the first video through a cosine similarity algorithm or Euclidean distance formula, or other similarity algorithms. No specific restrictions are imposed here. The calculated similarity can then be used as the facial matching degree between the original face image of the target object and the face images of each character in the face image of the first video.

[0146] Furthermore, after obtaining the facial matching degree between the original face image and the face images of each character in the first video, it can be understood that the higher the facial matching degree, the more similar the original face image of the target object is to the face image of the character object. Therefore, the face image of the character corresponding to the highest facial matching degree can be determined as the target character face image.

[0147] Optionally, in the above Figure 5 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application, such as... Figure 6 As shown, step S501 extracts facial feature information from the original face image, including:

[0148] In step S601, facial features are extracted from the original face image to obtain the original facial feature map;

[0149] In step S602, facial feature points are extracted from the original face image, and the extracted facial feature points are connected to obtain a facial contour feature map;

[0150] In step S603, the original facial feature map and the facial contour feature map are superimposed to obtain the target facial feature map.

[0151] In step S604, the facial features of the target face are decomposed to obtain facial feature information.

[0152] In this embodiment, after obtaining the original face image of the target object, the original face image can be extracted by the facial feature information extraction module to obtain the original facial feature map. At the same time, facial feature points can also be extracted from the original face image to obtain the feature points corresponding to the facial features on the original face image. These feature points are connected to obtain the facial contour feature map. Then, the original facial feature map and the facial contour feature map can be superimposed and optimized to obtain a feature-rich target facial feature map. Then, based on the feature-rich target facial feature map, facial feature decomposition processing is performed to obtain detailed and accurate facial feature information, which can improve the accuracy of recommending the target character's makeup scheme to a certain extent.

[0153] Specifically, such as Figure 13As shown, after obtaining the original face image of the target object, the original face image can be extracted by the facial feature information extraction module to obtain the original facial feature map. Specifically, the original face image can be processed by several fully connected layers or pooling layers in the convolutional neural network (CNN) to obtain the original facial feature map. Alternatively, the original face image can be extracted by other deep neural network frameworks such as SSD convolutional neural networks. No specific restrictions are made here.

[0154] Furthermore, facial feature points can be extracted from the acquired original face image, and a facial contour feature map can be obtained by connecting the extracted facial feature points. Specifically, the ASM algorithm can be used to detect anchor points of facial features such as facial contours, eyebrows, eyes, nose, or mouth from the face in the original face image. Then, the detected anchor points can be connected to form linear contours, such as facial contours and facial feature contours, to display the area of ​​facial features and obtain a facial contour feature map.

[0155] Furthermore, after obtaining the original facial feature map and facial contour feature map, the facial contour feature map containing linear contour anchor points can be aligned and superimposed with the original facial feature map containing the original human face features through scaling, transformation and translation operations. This can optimize pixels, reduce the influence of cluttered pixels, and improve image quality, resulting in a high-quality, information-rich target facial feature map. Then, the obtained target facial feature map can be processed by facial feature decomposition to obtain detailed and accurate facial feature information, such as features like almond-shaped eyes, double eyelids, high and low nose bridges, thick and thin lips, and thick and thin eyebrows.

[0156] Optionally, in the above Figure 5 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application, such as... Figure 7 As shown, skin condition information includes the oiliness, roughness, hemoglobin concentration, and melanin concentration of facial skin.

[0157] Step S501 extracts skin state information from the original face image, including:

[0158] In step S701, the original face image is subjected to intrinsic image decomposition processing to obtain a specular intrinsic layer, a diffuse intrinsic layer, and a skin color intrinsic layer.

[0159] In step S702, the oiliness of the facial skin is determined based on the ratio of the highlight intrinsic layer to the diffuse intrinsic layer.

[0160] In step S703, the roughness is determined based on the gradient value of the diffuse intrinsic layer;

[0161] In step S704, the pigment concentration of the intrinsic skin color layer is identified to obtain the hemoglobin concentration and melanin concentration.

[0162] In this embodiment, after obtaining the original facial image of the target object, intrinsic image decomposition processing can be performed on the original facial image to obtain the highlight intrinsic layer, diffuse intrinsic layer, and skin color intrinsic layer. Then, the oiliness of the facial skin can be determined based on the ratio between the highlight intrinsic layer and the diffuse intrinsic layer, and the roughness of the facial skin can be determined based on the gradient value of the diffuse intrinsic layer. Pigment concentration recognition can also be performed on the skin color intrinsic layer to identify the hemoglobin concentration and melanin concentration of the facial skin. By obtaining the ratio values ​​between the highlight intrinsic layer, diffuse intrinsic layer, and skin color intrinsic layer through decomposition, the skin condition information such as oiliness, roughness, hemoglobin concentration, and melanin concentration of the facial skin can be obtained more accurately, thereby improving the accuracy of the recommended makeup scheme to a certain extent.

[0163] The highlight intrinsic layer represents skin brightness; for example, the whiter the skin, the higher the brightness, or the oilier the skin, the higher the brightness. The diffuse intrinsic layer represents the skin's true color, which can be understood as the color of light reflected by the skin itself. For example, the color of objects in nature, such as grass, is green because the grass reflects the green light from sunlight to our eyes. The skin color intrinsic layer represents the skin color reflected by pigments such as melanin, proheme, and carotene in the epidermis.

[0164] Specifically, such as Figure 13 As shown, after obtaining the original face image of the target object, computer vision technology, such as Retinex theory, SIRFS algorithm and geometric structure as constraints, can be used to distinguish the material and color of the face itself, as well as the color and intensity distribution of the light source. The reflectance eigenmap corresponding to the material and color factors can be obtained, as well as the brightness eigenmap corresponding to the color and intensity distribution of the light source.

[0165] Furthermore, the reflectance intrinsic map and luminance intrinsic map can be calibrated by illumination using a color consistency algorithm. Then, using face intrinsic image decomposition technology, three intrinsic layers related to the skin color and texture of the target object can be decomposed from the reflectance intrinsic map and luminance intrinsic map: the highlight intrinsic layer, the diffuse intrinsic layer, and the skin color intrinsic layer.

[0166] Furthermore, after obtaining the highlight intrinsic layer, diffuse intrinsic layer, and skin color intrinsic layer, the oiliness of the skin can be determined by the ratio between the highlight intrinsic layer and the diffuse intrinsic layer. It can be understood that the higher the ratio of the highlight intrinsic layer to the diffuse intrinsic layer, the more oily the target's skin is, and conversely, the lower the ratio of the highlight intrinsic layer to the diffuse intrinsic layer, the more dry the target's skin is. Similarly, a ratio close to the median indicates that the target's skin is more combination-type.

[0167] Furthermore, since the diffuse intrinsic layer is used to represent the color of light reflected by the skin itself, the roughness of the skin can be expressed through the gradient information of the diffuse intrinsic layer itself. It can be understood that the larger the gradient of the diffuse intrinsic layer itself, the rougher the skin of the target object is, and conversely, the smaller the gradient of the diffuse intrinsic layer itself, the smoother the skin of the target object is.

[0168] Furthermore, since the skin color of the human epidermis is reflected by pigments such as melanin, proheme, and carotene, and the skin color of the face is mainly determined by the concentration of hemoglobin and melanin in skin cells, the pigment concentration of the intrinsic skin color layer can be identified by the dual-pigment linear model in skin biology to obtain the hemoglobin concentration and melanin concentration corresponding to the intrinsic skin color layer.

[0169] Optionally, in the above Figure 7 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application, such as... Figure 8 As shown, after extracting skin state information from the original face image in step S501, the method further includes:

[0170] In step S801, the database is read, and a skin quality report generation template is obtained from the database;

[0171] In step S802, the first skin care product category corresponding to the oiliness of the facial skin is obtained according to the preset correspondence between oiliness and skin care product category; the second skin care product category corresponding to roughness is obtained according to the preset correspondence between roughness and skin care product data; and the third skin care product category corresponding to pigment concentration is obtained according to the preset correspondence between hemoglobin concentration and melanin concentration and skin care product data.

[0172] In step S803, the intersection of the first skin care product category, the second skin care product category, and the third skin care product category is taken as the target skin care product set;

[0173] In step S804, the skin condition information is used to generate a skin condition report corresponding to the target object according to the report generation template;

[0174] In step S805, a skin condition report and a set of target skincare products are pushed to the target object.

[0175] In this embodiment, after extracting skin condition information from the original facial image, a database can be read to obtain a skin condition report generation template, a preset correspondence between oiliness and skincare product categories, a preset correspondence between roughness and skincare product data, and a preset correspondence between hemoglobin concentration and melanin concentration and skincare product data. Based on the obtained correspondence, the first skincare product category corresponding to oiliness, the second skincare product category corresponding to roughness, and the third skincare product category corresponding to pigment concentration can be quickly indexed. Then, the intersection of the first, second, and third skincare product categories can be used as the target skincare product set. The skin condition information is then used to generate a skin condition report corresponding to the target object according to the report generation template, and the skin condition report and the target skincare product set are pushed to the target object. By comprehensively considering the target skincare product set applicable to the four dimensions of facial skin oiliness, roughness, hemoglobin concentration, and melanin concentration, suitable skincare products can be better recommended to the target object.

[0176] The first category of skincare products refers to products that improve or maintain the skin's texture, such as oil-controlling cleansers, toners, and creams suitable for oily skin. The second category refers to products that smooth the skin or improve its texture, such as eye creams, eye serums, or anti-wrinkle neck creams. The third category refers to products that reduce dullness or brighten skin tone, such as whitening serums, sunscreens, or whitening face creams.

[0177] Specifically, such as Figure 13 As shown, the database can be read to obtain the corresponding relationships between preset oiliness and skincare product categories, preset roughness and skincare product data, and preset hemoglobin concentration and preset melanin concentration and skincare product data. Then, based on the correspondence between preset oiliness and skincare product categories, skincare products that match the oiliness of the target object (i.e., the first skincare product category) can be quickly indexed. Similarly, based on the correspondence between preset roughness and skincare product categories, skincare products that match the roughness of the target object (i.e., the second skincare product category) can be quickly indexed. Similarly, based on the correspondence between preset hemoglobin concentration and preset melanin concentration, skincare products that match the hemoglobin concentration and preset melanin concentration of the target object (i.e., the third skincare product category) can be quickly indexed.

[0178] Furthermore, the intersection of the first, second, and third skincare product categories can be calculated, and the intersection can be used as the target skincare product set. Alternatively, clustering or community discovery operations can be performed on the first, second, and third skincare product categories to identify products that are related to all three categories as target skincare products. The set of these products is the target skincare product set. Other classification operations can also be used to obtain the target skincare product categories, and no specific restrictions are imposed here.

[0179] Furthermore, to help the target audience understand their current skin condition more accurately and clearly, a skin condition report generation template can be retrieved from the database. This template contains a report style and generation rules. The skin condition information can then be used to generate a skin condition report for the target audience according to the template. Finally, the skin condition report and a set of target skincare products can be pushed to the target audience.

[0180] Understandably, skin condition information is primarily expressed through four dimensions: melanin concentration, hemoglobin concentration, roughness, and oiliness. Generally, melanin mainly affects skin color, which is often the primary consideration when choosing skincare products. Oiliness is a secondary factor; hemoglobin concentration, mainly manifested as varying degrees of fair and rosy skin, is usually the third factor; and skin roughness is the last factor. For example, assuming an ideal ratio of 4:2:3:1, a quadrilateral of this ratio can be used to filter a set of target skincare products from a wide range of options.

[0181] Optionally, in the above Figure 2 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application, such as... Figure 9 As shown, step S104 determines the target character's makeup scheme from several character makeup schemes based on the makeup matching degree, and pushes the target character's makeup scheme, including:

[0182] In step S901, the facial matching degree between the target character's face image and the original face image is weighted and summed with the makeup matching degree corresponding to each character's makeup scheme to obtain the makeup score corresponding to each character's makeup scheme.

[0183] In step S902, based on the makeup score, several character makeup schemes are filtered to obtain the target character makeup scheme and push it out.

[0184] In this embodiment, after obtaining the facial matching degree and makeup matching degree, the facial matching degree between the target character's face image and the original face image is weighted and summed with the makeup matching degree corresponding to each character's makeup scheme to obtain the makeup score corresponding to each character's makeup scheme. Then, several character makeup schemes can be filtered according to the size of the makeup score to obtain the target character makeup scheme suitable for the target object. By comprehensively considering the weight scores of facial similarity and makeup matching degree, the target character makeup scheme suitable for the target object can be better filtered.

[0185] Specifically, such as Figure 14 As shown, after obtaining facial similarity and makeup matching, given that makeup is highly dependent on face shape, the weight value of facial matching can be set to a larger weight value, and the weight value of makeup matching can be set to a smaller weight value. For example, the weight ratio between facial matching and makeup matching can usually be set to 7:3. Other ratios can also be set according to actual application needs, and no specific restrictions are made here.

[0186] Furthermore, based on the obtained weight values ​​of facial matching and makeup matching, a weighted sum can be calculated to obtain the makeup score corresponding to each character's makeup scheme. Then, the makeup scores corresponding to each character's makeup scheme can be compared pairwise to obtain the makeup scheme corresponding to the character with the highest makeup score as the target character's makeup scheme. Alternatively, the makeup scores corresponding to each character's makeup scheme can be sorted in descending order to obtain a makeup scheme recommendation set from largest to smallest. The character's makeup scheme corresponding to an element in the set can be prioritized and recommended to the target object. Makeup recommendations can also be made to the target object sequentially according to the order in the makeup scheme recommendation set. Other recommendation methods can also be used, without specific restrictions here.

[0187] Optionally, in the above Figure 2 Based on the corresponding embodiments, in another optional embodiment of the makeup scheme recommendation method provided in this application, such as... Figure 10 As shown, after step S104 determines the target character's makeup scheme from several character makeup schemes based on the makeup matching degree and pushes the target character's makeup scheme, the method further includes:

[0188] In step S1001, the face of the target object is scanned to obtain the current face scan image;

[0189] In step S1002, based on the unmade-up face image of the target object, the currently made-up area is extracted from the current face scan image;

[0190] In step S1003, the currently applied makeup area is compared with the standard makeup area of ​​the face image in the target character's makeup scheme to obtain the comparison result;

[0191] In step S1004, if the comparison result is consistent, a prompt to proceed to the next stage of makeup is sent to the target object;

[0192] In step S1005, if the comparison result is inconsistent, a prompt is sent to the target object indicating that the currently applied makeup area is inconsistent with the target character's makeup scheme.

[0193] In this embodiment, after pushing the target character's makeup scheme to the target object, the target object's face can be scanned in real time to obtain the current face scan image. Based on the target object's unmade-up face image, the currently made-up area is extracted from the current face scan image. Then, the currently made-up area can be compared with the standard makeup area of ​​the face image in the target character's makeup scheme to obtain the comparison result. When the comparison result is consistent, a prompt to proceed to the next stage of makeup can be sent to the target object. When the comparison result is inconsistent, a prompt can be sent to the target object that the currently made-up area is inconsistent with the target character's makeup scheme. By comparing the made-up area of ​​the current face scan image obtained in real time with the standard makeup area of ​​the face image in the target character's makeup scheme, it is possible to determine whether the target object's makeup operation is incorrect. Based on the judgment result, corresponding prompts can be issued to the target object to help the target object better learn and complete the reference makeup in the target character's makeup scheme.

[0194] The current facial scan image refers to the image of the target subject's entire makeup process, captured in real-time by a scanning device such as a camera. The unmade-up facial image of the target subject refers to the image of their bare face (without makeup) uploaded or scanned before they begin applying makeup according to the target character's makeup scheme. The currently applied makeup area refers to the areas of the target subject that are currently covered in makeup, such as the eyes, cheeks, or lips. The standard makeup area in the target character's makeup scheme refers to the area in the target character's makeup image that has a complete makeup effect.

[0195] Specifically, in practical application scenarios, the makeup tutorial for the target character's makeup scheme can consist of three parts: breaking down the makeup application steps, real-time makeup effects, and makeup modification suggestions. After pushing the target character's makeup scheme to the target audience, the makeup scheme can be played through the client on the terminal device. The tutorial can guide the target audience on the makeup application steps, positions, and angles through broken-down animations. For example, when applying foundation and concealer, the tutorial can demonstrate the techniques of pressing, smearing, or brushing depending on whether the concealer is for acne scars, dark circles, or eye bags. When shaping eyebrows, the tutorial can highlight the positions of the brow head, brow peak, and brow tail, or the stray hairs that need to be removed. When applying highlighter and contouring, the local area can be outlined with a dotted circle, and so on.

[0196] Furthermore, during the makeup application process, an image of the target's unmade-up face can be acquired first. A scanning device, such as a camera, can then capture the target's face scan image in real time throughout the entire makeup process, providing real-time feedback to the target and displaying the makeup effect. The current face scan image can then be compared with the target's unmade-up face image to determine the currently applied makeup area. Alternatively, a makeup recognition model can be used to identify the makeup in the current face scan image to determine the currently applied makeup area. Other methods can also be employed, without specific limitations.

[0197] Furthermore, after obtaining the currently applied makeup area, it can be compared one by one with the standard makeup area of ​​the face image in the target character's makeup scheme to obtain the comparison result of each applied makeup area on the face. If the comparison results of each applied makeup area are consistent, it can be understood that the currently applied makeup area meets the standard of the makeup effect in the target character's makeup scheme, and a prompt to proceed to the next stage of makeup can be sent to the target. Conversely, if the comparison results of any applied makeup area are inconsistent, it can be understood that the inconsistent applied makeup area does not meet the standard of the makeup effect in the target character's makeup scheme, and a prompt can be sent to the target that the currently applied makeup area is inconsistent with the target character's makeup scheme, and the inconsistent applied makeup area can be highlighted. The prompt can be made through a sidebar pop-up toast reminder and suggestion, or other forms, which are not specifically limited here.

[0198] The recommended apparatus for the makeup scheme in this application is described in detail below. Please refer to [link / reference]. Figure 15 , Figure 15 This is a schematic diagram of one embodiment of the makeup scheme recommendation device in this application. The makeup scheme recommendation device 20 includes:

[0199] The acquisition unit 201 is used to acquire the original face image of the target object in response to a first instruction triggered for the current video playback interface;

[0200] The acquisition unit 201 is also used to acquire a target character face image that matches the original face image, wherein the target character face image is derived from the character in the first video played on the current video playback interface;

[0201] Processing unit 202 is used to match the original face image with several character makeup schemes corresponding to the target character face image to obtain the makeup matching degree;

[0202] The determination unit 203 is used to determine the target character's makeup scheme from several character makeup schemes based on the makeup matching degree, and to push the target character's makeup scheme.

[0203] Optionally, in the above Figure 15 Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application,

[0204] The acquisition unit is also used to acquire the basic object face image, the character makeup sample image, and the basic attribute features of the character corresponding to the character in the character makeup sample image. The character makeup sample image is any frame face image of each film and television character extracted from the film and television works, and the character makeup sample image has a corresponding makeup label.

[0205] The processing unit is also used to extract the facial contour features of the base object from the base object's facial image, and to extract the character's makeup features and the character's facial contour features from the makeup sample image.

[0206] The processing unit is also used to input the basic object's facial contour features, the character's basic attribute features, the character's makeup features, and the character's facial contour features into the makeup prediction model, and output the makeup prediction probability through the makeup prediction model.

[0207] The processing unit is also used to update the model parameters of the makeup prediction model based on the makeup prediction probability and the makeup label;

[0208] The processing unit can be used to: input the original face image into the makeup prediction model, output the makeup prediction probability corresponding to each character's makeup scheme through the makeup prediction model, and determine the makeup matching degree between the original face image and each character's makeup scheme based on the makeup prediction probability.

[0209] Optionally, in the above Figure 15 Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application, the target character makeup scheme includes makeup tutorial information and makeup product information;

[0210] Makeup tutorial information includes several makeup steps, and each makeup step includes at least one of the following: instructions for using the makeup products and a makeup effect image.

[0211] Makeup product information includes base makeup product information and color makeup product information, which appears in makeup tutorial information.

[0212] Optionally, in the above Figure 15 Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application, the acquisition unit 201 can specifically be used for:

[0213] Perform a facial scan on the target object to obtain the original face image; or

[0214] The system identifies the face photos uploaded by the target object to obtain the original face image.

[0215] Optionally, in the above Figure 15 Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application, the acquisition unit 201 can specifically be used for:

[0216] Extract facial feature information from the original face image, and extract skin state information from the original face image;

[0217] Based on facial feature information and skin condition information, calculate the facial matching degree between the original face image and the face images of each character in the first video played on the current video playback interface.

[0218] Based on facial matching accuracy, the target character's facial image is determined from the facial images of each character in the first video.

[0219] Optionally, in the above Figure 15 Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application, the acquisition unit 201 can specifically be used for:

[0220] Facial feature extraction is performed on the original face image to obtain the original facial feature map;

[0221] Facial feature points are extracted from the original face image, and the extracted facial feature points are connected to obtain a facial contour feature map;

[0222] The original facial feature map and the facial contour feature map are superimposed to obtain the target facial feature map.

[0223] The facial features of the target face are decomposed to obtain facial feature information.

[0224] Optionally, in the above Figure 15Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application, the acquisition unit 201 can specifically be used for:

[0225] The original face image is subjected to intrinsic image decomposition processing to obtain the specular intrinsic layer, diffuse intrinsic layer and skin color intrinsic layer;

[0226] The oiliness of facial skin is determined by the ratio of the highlight intrinsic layer to the diffuse intrinsic layer.

[0227] The roughness is determined based on the gradient value of the diffuse intrinsic layer;

[0228] Pigment concentration was identified in the intrinsic layer of skin color to obtain hemoglobin concentration and melanin concentration.

[0229] Optionally, in the above Figure 15 Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application,

[0230] The acquisition unit 201 is also used to read the database and retrieve the skin quality report generation template from the database;

[0231] The acquisition unit 201 is also used to acquire the first skin care product category corresponding to the oiliness of the human face skin according to the preset correspondence between oiliness and skin care product category, and to acquire the second skin care product category corresponding to roughness according to the preset correspondence between roughness and skin care product data, and to acquire the third skin care product category corresponding to pigment concentration according to the preset correspondence between hemoglobin concentration and preset melanin concentration and skin care product data.

[0232] The processing unit 202 is also used to take the intersection of the first skin care category, the second skin care category and the third skin care category as the target skin care product set;

[0233] The processing unit 202 is also used to generate a skin condition report corresponding to the target object based on the skin condition information according to the report generation template;

[0234] The processing unit 202 is also used to push skin condition reports and target skin care product sets to the target object.

[0235] Optionally, in the above Figure 15 Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application, the determining unit 203 can specifically be used for:

[0236] The facial matching degree between the target character's face image and the original face image is weighted and summed with the makeup matching degree corresponding to each character's makeup scheme to obtain the makeup score corresponding to each character's makeup scheme.

[0237] Based on the makeup score, several character makeup schemes are filtered to obtain the target character's makeup scheme and push it to the user.

[0238] Optionally, in the above Figure 15 Based on the corresponding embodiments, in another embodiment of the makeup scheme recommendation device provided in this application,

[0239] The acquisition unit 201 is also used to scan the face of the target object and acquire the current face scan image;

[0240] The acquisition unit 201 is also used to extract the currently made-up area from the current face scan image based on the unmade-up face image of the target object;

[0241] The processing unit 202 is also used to compare the currently applied makeup area with the standard makeup area of ​​the face image in the target character's makeup scheme to obtain a comparison result;

[0242] The processing unit 202 is also used to send a prompt to the target object to proceed to the next stage of makeup if the comparison result is consistent;

[0243] The processing unit 202 is also used to send a prompt to the target object that the currently applied makeup area is inconsistent with the target character's makeup scheme if the comparison result is inconsistent.

[0244] This application also provides a schematic diagram of another computer device, such as... Figure 16 As shown, Figure 16 This is a schematic diagram of a computer device structure provided in an embodiment of this application. The computer device 300 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors) and a memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) for storing application programs 331 or data 332. The memory 320 and storage media 330 can be temporary or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the computer device 300. Furthermore, the CPU 310 may be configured to communicate with the storage media 330 and execute the series of instruction operations in the storage media 330 on the computer device 300.

[0245] Computer device 300 may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 333, such as Windows Server. TM Mac OS X TM Unix TM Linux TM FreeBSD TM etc.

[0246] The aforementioned computer device 300 is also used to perform, for example Figures 2 to 10 The steps in the corresponding embodiments.

[0247] Another aspect of this application provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform actions such as... Figures 2 to 10 The steps in the method described in the illustrated embodiment.

[0248] Another aspect of this application provides a computer program product containing instructions that, when run on a computer or processor, cause the computer or processor to perform actions such as Figures 2 to 10 The steps in the method described in the illustrated embodiment.

[0249] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0250] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0251] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0252] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0253] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for recommending a makeup look, characterized in that, include: In response to the first command triggered by the current video playback interface, the original face image of the target object is obtained; Obtain a target character face image that matches the original face image, wherein the target character face image originates from a character in the first video played on the current video playback interface; The original face image is matched with several character makeup schemes corresponding to the target character face image to obtain the makeup matching degree; the makeup matching degree is obtained based on a makeup prediction model. Based on the makeup matching degree, a target character makeup scheme is determined from the plurality of character makeup schemes, and the target character makeup scheme is pushed out. The training process of the makeup prediction model includes: acquiring a basic object face image, a character makeup sample image, and the basic attribute features of the character corresponding to the character in the character makeup sample image. The character makeup sample image is any frame face image of each film and television character extracted from the film and television works, and the character makeup sample image corresponds to a makeup label. Extract the basic object's facial contour features from the basic object's facial image, and extract the character's makeup features and character's facial contour features from the makeup sample image; The basic object's facial contour features, the character's basic attribute features, the character's makeup features, and the character's facial contour features are input into the makeup prediction model, and the makeup prediction model outputs the makeup prediction probability. The model parameters of the makeup prediction model are updated using the predicted makeup probability and the makeup label.

2. The method according to claim 1, characterized in that, The step of matching the original facial image with several character makeup schemes corresponding to the target character's facial image to obtain a makeup matching degree includes: The original face image is input into the makeup prediction model, and the makeup prediction model outputs the makeup prediction probability corresponding to each character's makeup scheme. Based on the makeup prediction probability, the makeup matching degree between the original face image and each character's makeup scheme is determined.

3. The method according to claim 1, characterized in that, The target character's makeup plan includes makeup tutorial information and makeup product information; The makeup tutorial information includes several makeup application steps, and each makeup application step includes at least one of the following: instructions for using the makeup product and a makeup effect image. The makeup product information includes base makeup product information and color makeup product information, which appears in the makeup tutorial information.

4. The method according to claim 1, characterized in that, The process of obtaining the original facial image of the target object includes: Perform a facial scan on the target object to obtain the original face image; or The original face image is obtained by recognizing the face photo uploaded by the target object.

5. The method according to claim 1, characterized in that, The step of obtaining the target character's face image that matches the original face image includes: Facial feature information is extracted from the original face image, and skin condition information is extracted from the original face image; Based on the facial feature information and the skin condition information, calculate the facial matching degree between the original face image and the face images of each character in the first video played on the current video playback interface; Based on the facial matching degree, the target character's facial image is determined from the facial images of each character in the first video.

6. The method according to claim 5, characterized in that, The step of extracting facial feature information from the original face image includes: Facial feature extraction is performed on the original face image to obtain the original facial feature map; Facial feature points are extracted from the original face image, and a facial contour feature map is obtained by connecting the extracted facial feature points. The original facial feature map and the facial contour feature map are superimposed to obtain the target facial feature map. The facial features of the target face are decomposed into facial features to obtain the facial feature information.

7. The method according to claim 5, characterized in that, The skin condition information includes the oiliness, roughness, hemoglobin concentration, and melanin concentration of the facial skin; Extracting skin state information from the original face image includes: The original face image is subjected to intrinsic image decomposition processing to obtain a specular intrinsic layer, a diffuse intrinsic layer, and a skin color intrinsic layer; The oiliness of the facial skin is determined based on the ratio of the highlight intrinsic layer to the diffuse intrinsic layer. The roughness is determined based on the gradient value of the diffuse intrinsic layer; The pigment concentration of the intrinsic skin color layer is identified to obtain the hemoglobin concentration and the melanin concentration.

8. The method according to claim 7, characterized in that, After extracting skin state information from the original face image, the method further includes: Read the database and retrieve the skin quality report generation template from the database; Based on the preset correspondence between oiliness and skincare product categories, the first skincare product category corresponding to the oiliness of the facial skin is obtained; based on the preset correspondence between roughness and skincare product data, the second skincare product category corresponding to the roughness is obtained; and based on the preset correspondence between hemoglobin concentration and melanin concentration and skincare product data, the third skincare product category corresponding to the pigment concentration is obtained. The intersection of the first skincare category, the second skincare category, and the third skincare category is taken as the target skincare product set; The skin condition information is used to generate a skin condition report corresponding to the target object according to the report generation template. The skin condition report and the set of target skincare products are pushed to the target object.

9. The method according to claim 5, characterized in that, The step of determining a target character's makeup scheme from the plurality of character makeup schemes based on the makeup matching degree, and pushing the target character's makeup scheme, includes: The facial matching degree between the target character's facial image and the original facial image is weighted and summed with the makeup matching degree corresponding to each character's makeup scheme to obtain the makeup score corresponding to each character's makeup scheme. Based on the makeup score, the makeup schemes for the various characters are filtered to obtain the makeup scheme for the target character and then pushed to the user.

10. The method according to claim 1, characterized in that, After determining the target character's makeup scheme from the plurality of character makeup schemes based on the makeup matching degree, and pushing the target character's makeup scheme, the method further includes: Scan the face of the target object to obtain the current face scan image; Based on the unmade-up face image of the target object, extract the currently made-up area from the current face scan image; The currently applied makeup area is compared with the standard makeup area of ​​the face image in the target character's makeup scheme to obtain the comparison result; If the comparison results are consistent, a prompt to proceed to the next stage of makeup is sent to the target object; If the comparison result is inconsistent, a prompt is sent to the target object indicating that the currently applied makeup area is inconsistent with the target character's makeup scheme.

11. A device for recommending makeup schemes, characterized in that, include: The acquisition unit is used to acquire the original face image of the target object in response to a first instruction triggered for the current video playback interface; The acquisition unit is further configured to acquire a target character face image that matches the original face image, wherein the target character face image originates from a character in the first video played on the current video playback interface; The processing unit is used to match the original face image with several character makeup schemes corresponding to the target character face image to obtain a makeup matching degree; the makeup matching degree is obtained based on a makeup prediction model. The determining unit is used to determine the target character makeup scheme from the plurality of character makeup schemes based on the makeup matching degree, and to push the target character makeup scheme. The acquisition unit is also used to acquire a basic object face image, a character makeup sample image, and basic attribute features of the character corresponding to the character in the character makeup sample image. The character makeup sample image is any frame face image of each film and television character extracted from the film and television works, and the character makeup sample image has a corresponding makeup tag. The processing unit is also used to extract the basic object face contour features from the basic object face image, and to extract the character makeup features and character face contour features from the makeup sample image. The processing unit is also used to input the basic object's facial contour features, the character's basic attribute features, the character's makeup features, and the character's facial contour features into the makeup prediction model, and output the makeup prediction probability through the makeup prediction model. The processing unit is further configured to update the model parameters of the makeup prediction model using the makeup prediction probability and the makeup label.

12. The apparatus according to claim 11, characterized in that, The processing unit can be specifically used to: input the original face image into the makeup prediction model, output the makeup prediction probability corresponding to each character's makeup scheme through the makeup prediction model, and determine the makeup matching degree between the original face image and each character's makeup scheme based on the makeup prediction probability.

13. The apparatus according to claim 11, characterized in that, The target character's makeup scheme includes makeup tutorial information and makeup product information; the makeup tutorial information includes several makeup application steps, and each makeup application step includes at least one of the following: instructions for using the makeup product and a makeup effect image; The makeup product information includes base makeup product information and color makeup product information, which appears in the makeup tutorial information.

14. The apparatus according to claim 11, characterized in that, The acquisition unit can specifically be used for: Perform a facial scan on the target object to obtain the original face image; or The original face image is obtained by recognizing the face photo uploaded by the target object.

15. The apparatus according to claim 11, characterized in that, The acquisition unit can specifically be used for: Facial feature information is extracted from the original face image, and skin condition information is extracted from the original face image; Based on the facial feature information and the skin condition information, calculate the facial matching degree between the original face image and the face images of each character in the first video played on the current video playback interface; Based on the facial matching degree, the target character's facial image is determined from the facial images of each character in the first video.

16. The apparatus according to claim 15, characterized in that, The acquisition unit can specifically be used for: Facial feature extraction is performed on the original face image to obtain the original facial feature map; Facial feature points are extracted from the original face image, and a facial contour feature map is obtained by connecting the extracted facial feature points. The original facial feature map and the facial contour feature map are superimposed to obtain the target facial feature map. The facial features of the target face are decomposed into facial features to obtain the facial feature information.

17. The apparatus according to claim 15, characterized in that, The skin condition information includes the oiliness, roughness, hemoglobin concentration, and melanin concentration of the facial skin; the acquisition unit can specifically be used for: The original face image is subjected to intrinsic image decomposition processing to obtain a specular intrinsic layer, a diffuse intrinsic layer, and a skin color intrinsic layer; The oiliness of the facial skin is determined based on the ratio of the highlight intrinsic layer to the diffuse intrinsic layer. The roughness is determined based on the gradient value of the diffuse intrinsic layer; The pigment concentration of the intrinsic skin color layer is identified to obtain the hemoglobin concentration and the melanin concentration.

18. The apparatus according to claim 17, characterized in that, The acquisition unit is also used to read the database and obtain a skin quality report generation template from the database; The acquisition unit is further configured to acquire the first skin care product category corresponding to the oiliness of the facial skin according to the preset correspondence between oiliness and skin care product category, acquire the second skin care product category corresponding to the roughness according to the preset correspondence between roughness and skin care product data, and acquire the third skin care product category corresponding to the pigment concentration according to the preset correspondence between hemoglobin concentration and melanin concentration and skin care product data. The processing unit is further configured to take the intersection of the first skin care product category, the second skin care product category, and the third skin care product category as the target skin care product set; The processing unit is also used to generate a skin condition report corresponding to the target object based on the skin condition information according to the report generation template; The processing unit is also used to push the skin condition report and the target skin care product set to the target object.

19. The apparatus according to claim 15, characterized in that, The determining unit can specifically be used for: The facial matching degree between the target character's facial image and the original facial image is weighted and summed with the makeup matching degree corresponding to each character's makeup scheme to obtain the makeup score corresponding to each character's makeup scheme. Based on the makeup score, the makeup schemes for the various characters are filtered to obtain the makeup scheme for the target character and then pushed to the user.

20. The apparatus according to claim 11, characterized in that, The acquisition unit is also used to scan the face of the target object to acquire the current face scan image; The acquisition unit is further configured to extract the currently made-up area from the current face scan image based on the unmade-up face image of the target object; The processing unit is also used to compare the currently applied makeup area with the standard makeup area of ​​the face image in the target character's makeup scheme to obtain a comparison result; The processing unit is also configured to send a prompt to the target object to proceed to the next stage of makeup if the comparison result is consistent. The processing unit is further configured to send a prompt to the target object that the currently applied makeup area is inconsistent with the target character's makeup scheme if the comparison result is inconsistent.

21. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 10.

22. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 10.

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

Images