Fitness teaching method and system based on virtual human

Abstract

The invention relates to a body building teaching method and system based on a virtual human, and belongs to the field of three-dimensional modeling, and the body building teaching method based on the virtual human comprises the steps: obtaining a figure image, generating a 3D virtual human model according to the figure image through an artificial intelligence large model, generating a virtual human teaching animation according to the 3D virtual human model and a preset body building skeleton animation, and displaying the virtual human teaching animation to the target user so as to indicate the target user to carry out action following learning according to the virtual human teaching animation. According to the invention, the problem that the recorded teaching video cannot comprehensively and accurately convey the teaching content is solved. According to the virtual human teaching animation, the teaching process of the 3D virtual human model can be displayed from various angles, and the teaching content can be more comprehensively and accurately conveyed.

Description

Technical Field

[0001] The present application relates to the field of three-dimensional modeling, and in particular to a fitness teaching method and system based on virtual humans. Background Art

[0002] As people pay more attention to health and body shape, more and more people start to exercise. Existing fitness teaching is usually carried out through personal guidance by fitness coaches or watching recorded teaching videos. However, some fitness movements are complicated, and recorded teaching videos can only show the movements from one angle, which cannot fully and accurately convey the teaching content, resulting in learners being unable to accurately understand the movements; and the way coaches teach fitness in person requires going to the gym, which is time-consuming and laborious.

[0003] Currently, no effective solution has been proposed for the problem that the teaching videos recorded in related technologies cannot fully and accurately convey the teaching content. Summary of the invention

[0004] The embodiments of the present application provide a fitness teaching method, system, electronic device and storage medium based on a virtual human, so as to at least solve the problem in the related art that the recorded teaching videos cannot fully and accurately convey the teaching content.

[0005] In a first aspect, an embodiment of the present application provides a fitness teaching method based on a virtual person, the method comprising:

[0006] Acquire a character image, and generate a 3D virtual human model based on the character image through an artificial intelligence model;

[0007] Generate a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation;

[0008] The virtual human teaching animation is displayed to the target user to instruct the target user to follow the movements of the virtual human teaching animation.

[0009] In some embodiments, presenting the virtual human teaching animation to the target user includes:

[0010] Displaying the virtual human teaching screen through a projection screen or an AR device, wherein the virtual human teaching screen is a 3D screen;

[0011] According to the angle adjustment instruction of the target user, the display angle of the 3D virtual human in the virtual human teaching screen is adjusted.

[0012] In some embodiments, generating a 3D virtual human model according to the character image by using an artificial intelligence big model includes:

[0013] Performing multimodal learning through the artificial intelligence big model, analyzing and acquiring visual features in the character image, and associating the visual features with corresponding 3D spatial information to obtain associated information;

[0014] Extracting image features from the person image based on the deep neural network in the artificial intelligence big model;

[0015] The person in the person image is three-dimensionally reconstructed according to the association information and the image features to obtain the 3D virtual human model.

[0016] In some embodiments, the method further comprises:

[0017] When the target user is performing action following, identifying the fitness action of the target user to obtain an action recognition result;

[0018] The fitness movements of the target user are evaluated according to the movement recognition results and the virtual human teaching animation.

[0019] In some embodiments, the evaluating the fitness action of the target user according to the action recognition result and the virtual human teaching animation includes:

[0020] Obtaining first position information of preset body skeleton key points of the target user according to the action recognition result;

[0021] Acquire second position information of the preset body skeleton key point of the 3D virtual human model, wherein the 3D virtual human model is the 3D virtual human model in the virtual human teaching animation;

[0022] The first location information and the second location information are compared to evaluate the fitness action of the target user.

[0023] In some embodiments, generating a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation includes:

[0024] In the case where the character image is a character image of the target user, obtaining the physical features of the target user from the character image, and obtaining the historical learning record of the target user;

[0025] According to the physical features and / or the historical learning records, matching a target fitness course, and obtaining a preset fitness skeleton animation corresponding to the target fitness course;

[0026] A virtual human teaching animation is generated according to the 3D virtual human model and the preset fitness skeleton animation of the target fitness course.

[0027] In some embodiments, the method further comprises:

[0028] The scene picture of the target user performing action following is collected and saved, wherein the scene picture includes the target user and the virtual human teaching animation.

[0029] In a second aspect, the embodiment of the present application provides a fitness teaching system based on a virtual person, the system comprising: a model building module, an animation generation module and a display module, wherein:

[0030] The model building module is used to obtain a character image and generate a 3D virtual human model based on the character image through an artificial intelligence large model;

[0031] The animation generation module is used to generate a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation;

[0032] The display module is used to display the virtual human teaching animation to the target user, so as to instruct the target user to follow the movements of the virtual human teaching animation.

[0033] In a third aspect, an embodiment of the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the virtual human-based fitness teaching method as described in the first aspect above is implemented.

[0034] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the fitness teaching method based on a virtual human as described in the first aspect above.

[0035] Compared with the related art, the fitness teaching method based on virtual human provided by the embodiment of the present application obtains the character image, generates a 3D virtual human model according to the character image through the artificial intelligence big model, generates a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation, and displays the virtual human teaching animation to the target user to instruct the target user to follow the movements according to the virtual human teaching animation, thus solving the problem that the recorded teaching video cannot fully and accurately convey the teaching content. The virtual human teaching animation can display the teaching process of the 3D virtual human model from various angles, and can convey the teaching content more comprehensively and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0037] Figure 1 is a schematic diagram of an application environment of a fitness teaching method based on a virtual person according to an embodiment of the present application;

[0038] Figure 2 is a flow chart of a fitness teaching method based on a virtual person according to an embodiment of the present application;

[0039] Figure 3 is a schematic diagram of a body bone position marker according to an embodiment of the present application;

[0040] Figure 4 is a structural block diagram of a fitness teaching system based on a virtual person according to an embodiment of the present application;

[0041] Figure 5 It is a schematic diagram of the internal structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0043] Obviously, the drawings described below are only some examples or embodiments of the present application. For ordinary technicians in this field, the present application can also be applied to other similar scenarios based on these drawings without creative work. In addition, it can also be understood that although the efforts made in this development process may be complicated and lengthy, for ordinary technicians in this field related to the content disclosed in this application, some changes in design, manufacturing or production based on the technical content disclosed in this application are just conventional technical means, and should not be understood as insufficient content disclosed in this application.

[0044] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0045] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantitative limitation, and may represent the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0046] Figure 1 is a schematic diagram of an application environment of a fitness teaching method based on a virtual person according to an embodiment of the present application. The fitness teaching method based on a virtual person provided by the present application can be applied to Figure 1 In the application environment shown. Figure 1 As shown, the server 101 communicates with the terminal 102 through the network. The server 101 obtains the character image, generates a 3D virtual human model based on the character image through the artificial intelligence big model, generates a virtual human teaching animation based on the 3D virtual human model and the preset fitness skeleton animation, and the server 101 sends the virtual human teaching animation to the terminal 102, and the terminal 102 displays the virtual human teaching animation to the target user to instruct the target user to follow the action according to the virtual human teaching animation. Among them, the terminal 102 includes but is not limited to an AR device and a projection device.

[0047] This embodiment provides a fitness teaching method based on a virtual human. Figure 2 is a flowchart of a fitness teaching method based on a virtual person according to an embodiment of the present application. Figure 2 As shown, the process includes the following steps:

[0048] Step S201, obtain a character image, and generate a 3D virtual human model based on the character image through an artificial intelligence large model.

[0049] It should be noted that the character image can be a character picture taken in real time by a photographic device, for example, an external camera device is set to obtain a full-body flat image or stereoscopic image of the user, or it can be an image picture of some favorite characters or cartoon characters uploaded by the user. While obtaining the user's image, the user's image and the virtual image picture are combined to generate a 3D virtual human model. The image of the 3D virtual human model is the same as the avatar of the character or cartoon character in the image picture, and the body movements of the 3D virtual human model are consistent with the user's movements.

[0050] It should also be noted that in addition to obtaining character images, the user's body movements can also be obtained through the sensor system on the user's body to generate a 3D virtual human model that follows the user's body movements in real time.

[0051] In some embodiments, generating a 3D virtual human model according to a human image by using an artificial intelligence large model in step S201 includes:

[0052] Step S2011, multimodal learning is performed through a large artificial intelligence model to analyze and obtain visual features in the character image, and the visual features are associated with corresponding 3D spatial information to obtain associated information.

[0053] The AI ​​big model performs multimodal learning and simultaneously processes data from different modalities, such as images and 3D structures, to understand the visual features in images of people and associate the visual features with the corresponding 3D spatial information.

[0054] Step S2012, extracting image features from the character image based on the deep neural network in the artificial intelligence large model.

[0055] Image features are extracted from human images using deep neural networks, especially convolutional neural networks (CNNs). The image features include but are not limited to surface color and texture, shape, contour and contextual relationship of the human figure.

[0056] Step S2013, performing three-dimensional reconstruction on the person in the person image according to the associated information and the image features to obtain a 3D virtual human model.

[0057] The large artificial intelligence model estimates the depth information of the character by learning a large number of annotated 2D-3D pairs and based on geometric reasoning techniques (such as depth estimation or stereo matching), thereby predicting the 3D geometric shape based on the 2D projection and reconstructing the character in the character image in three dimensions.

[0058] For example, a 3D virtual human model is reconstructed through Neural Radiance Fields (NeRF). NeRF learns the distribution of the scene through a continuum of multi-layer perceptrons (MLPs), which can synthesize images from different perspectives to derive implicit 3D structures.

[0059] In this embodiment, the artificial intelligence large model can also be combined with traditional graphics technology to reversely map the 2D character image to a possible 3D model through a reverse rendering process, and then adjust the model parameters through an optimization algorithm to match the input image to the greatest extent.

[0060] In some cases, generative models such as generative adversarial networks (GANs) or variational autoencoders (VAEs) can be used to progressively generate realistic 3D models that incorporate semantic and structural features learned from 2D images starting from a random seed or low-dimensional representation.

[0061] The process of generating a 3D model is often not a one-step process, but rather an iterative optimization and feedback loop to refine the model details until it achieves the best result under a certain metric (e.g., Chamfer distance, IoU, or perceptual loss function).

[0062] Through model rendering, the texture information in the original character image is mapped onto the surface of the 3D virtual human model to increase the realism and details of the model.

[0063] Continue to refer to Figure 2 , after generating the 3D virtual human model, execute step S202.

[0064] Step S202: Generate a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation.

[0065] Through the pre-made fitness skeleton animation, add bones to the 3D virtual human model, and then add animation controls to bind the preset skeleton animation to the 3D virtual human model. Optionally, the skeleton animation binding is realized through the main Unity development tool.

[0066] The 3D virtual human model is bound to fitness movements to obtain a virtual human teaching animation. Users can follow the teaching animation accurately and can also upload their favorite virtual human to accompany them in fitness.

[0067] It should be noted that multiple 3D virtual human models can be generated, and the generated virtual human models are displayed on the interface ui. The model is selected through user instructions and moved up, down, left, and right to adjust to a suitable position.

[0068] In some embodiments, step S202 specifically includes:

[0069] Step S2021, when the character image is a character image of a target user, obtain the physical features of the target user from the character image, and obtain the historical learning record of the target user.

[0070] Step S2022, matching the target fitness course according to the physical features and / or historical learning records, and obtaining the preset fitness skeleton animation corresponding to the target fitness course.

[0071] Step S2023, generating a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation of the target fitness course.

[0072] The most suitable fitness courses are matched to target users based on their physical characteristics and historical learning records. For example, through the big data intelligent recommendation algorithm, other courses that are commonly used and liked by users of the same type are recommended based on the fitness courses that the users often take.

[0073] Step S203: display the virtual human teaching animation to the target user to instruct the target user to follow the movements of the virtual human teaching animation.

[0074] In some embodiments, step S203 specifically includes:

[0075] Step S2031, displaying a virtual human teaching screen through a projection screen or an AR device, where the virtual human teaching screen is a 3D screen.

[0076] It should be noted that, in the case of a projection screen, the user's image can be obtained through the projector's camera or an external camera device to generate a 3D virtual human model, and the 3D virtual human model and virtual human teaching animation are displayed on the projection screen at the same time. The user can directly see the 3D virtual human model and the virtual human teaching animation on the projection screen, and the user can see his or her fitness effect more intuitively.

[0077] When the user's terminal is wearable glasses of augmented reality (AR) or virtual reality (VR), the user's character image can be obtained through the camera of the wearable glasses or an external camera device, and the user's body movements can also be obtained through the user's own wearable device, such as a posture sensor of the user's wearable device in the form of clothes. The character image and the user's body movements are combined, and a 3D simulated human model is generated through an artificial intelligence large model. The 3D virtual human model and the virtual human teaching animation are displayed on the display interface of the wearable glasses at the same time. The display interface can be a lens of an AR device or a display screen of a VR device. The user is not restricted by place and can exercise at any time while wearing wearable glasses.

[0078] Step S2032, adjusting the display angle of the 3D virtual human in the virtual human teaching screen according to the angle adjustment instruction of the target user.

[0079] In this embodiment, according to user instructions, the virtual human teaching animation can display the teaching process of the 3D virtual human model from various angles, and convey the teaching content more comprehensively and accurately.

[0080] Through the above steps, the fitness teaching method based on virtual human provided by the embodiment of the present application obtains the character image, generates a 3D virtual human model according to the character image through the artificial intelligence large model, generates a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation, and displays the virtual human teaching animation to the target user to instruct the target user to follow the movements according to the virtual human teaching animation, solving the problem that the recorded teaching video cannot fully and accurately convey the teaching content. The virtual human teaching animation can display the teaching process of the 3D virtual human model from all angles, and can convey the teaching content more comprehensively and accurately.

[0081] In some embodiments, the method further comprises:

[0082] Step S204, collecting and saving the scene picture of the target user following the action, where the scene picture includes the target user and the virtual human teaching animation.

[0083] The camera equipment pre-set in the environment is used to capture the scene images of the target user following the movements and learning, and to record the process of the user learning fitness with the 3D virtual person for commemoration, sharing and display. The user can also review his or her own learning process based on the scene images.

[0084] In some embodiments, the method further comprises:

[0085] Step S205, when the target user is following the movements, the fitness movements of the target user are recognized to obtain movement recognition results.

[0086] Step S206: Evaluate the fitness movements of the target user according to the movement recognition results and the virtual human teaching animation.

[0087] In some embodiments, step S206 specifically includes:

[0088] Step S2061, obtaining first position information of preset body skeleton key points of the target user according to the action recognition result.

[0089] Step S2062, obtaining second position information of preset body skeleton key points of the 3D virtual human model, where the 3D virtual human model is a 3D virtual human model in the virtual human teaching animation.

[0090] Step S2063: compare the first position information and the second position information to evaluate the fitness action of the target user.

[0091] Figure 3 is a schematic diagram of a body bone position marker according to an embodiment of the present application, such as Figure 3 As shown in the figure, 14 key points of the human skeleton are marked on the human body through the camera and human body recognition algorithm. The first coordinate system is constructed with the upper left boundary point of the video as the origin, the right as the positive direction of the X axis, and the downward as the positive direction of the Y axis, and the coordinates of these 14 points are obtained.

[0092] The target user's learning image and the virtual human teaching animation obtained in real time are respectively identified to obtain the first position information of the target user's body skeleton key points and the second position information of the 3D virtual human's limb coordinates corresponding to the body skeleton key points. The fitness movements of the target user are evaluated based on the first position information and the second position information of each body skeleton key point, as well as the distance between each body skeleton key point. In this embodiment, the limb coordinates of the 3D virtual human can be obtained in advance and saved in the cloud.

[0093] Optionally, the coordinate system is transformed with the center of the character as the coordinate origin, so as to more intuitively display the skeleton key points and their corresponding position deviations to the user.

[0094] Assume that the limb coordinate points of the 3D virtual person in the first coordinate system are A1(x1,y1), A2(x1,y1), ..., A13(x13,y14), A14(x14,y14). Create a new Cartesian coordinate system, and use the center point of A3, A6, A9, and A10 as the new origin. The origin of the obtained second coordinate system corresponds to the point with coordinates ((x3+x6+x9+x10) / 4, (y3+y6+y9+y10) / 4) in the first coordinate system.

[0095] Through coordinate transformation, the 3D virtual human is transformed from the first coordinate system to the second coordinate system, and the corresponding coordinates of the 14 limb points are transformed into B1(m1,n1), B2(m2,n2),..., B14(m14,n14). For the entire fitness course, the coach's standard movements are obtained every 2 seconds, and after converting them into the new coordinate system, the data is saved in the cloud together with the previous limb movement data for backup.

[0096] In this embodiment, corresponding sound effects can be set according to the standard degree of user actions, and sound effect rewards (such as praise voice) can be added through different matching levels to encourage users to learn. Check-in points can also be set, which can be exchanged for different badges for sharing, displaying, exchanging for fitness memberships, etc., to improve the user's enthusiasm for learning.

[0097] It should be noted that the steps shown in the above process or the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0098] This embodiment also provides a fitness teaching system based on a virtual person, which is used to implement the above embodiments and preferred implementations, and will not be repeated here. As used below, the terms "module", "unit", "subunit", etc. can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0099] Figure 4 is a structural block diagram of a fitness teaching system based on a virtual person according to an embodiment of the present application, such as Figure 4 As shown, the system includes: a model building module 41, an animation generation module 42 and a display module 43, wherein:

[0100] The model building module 41 is used to obtain a character image and generate a 3D virtual human model based on the character image through an artificial intelligence large model.

[0101] The animation generation module 42 is used to generate a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation.

[0102] The display module 43 is used to display the virtual human teaching animation to the target user, so as to instruct the target user to follow the movements according to the virtual human teaching animation.

[0103] In some embodiments, the display module 43 includes: a picture display module and an angle adjustment module.

[0104] The screen display module is used to display the virtual human teaching screen through a projection screen or AR device. The virtual human teaching screen is a 3D screen.

[0105] The angle adjustment module is used to adjust the display angle of the 3D virtual human in the virtual human teaching screen according to the angle adjustment instruction of the target user.

[0106] In some of the embodiments, the model building module 41 includes: an association module, a feature extraction module and a three-dimensional reconstruction module.

[0107] The association module is used to perform multimodal learning through a large artificial intelligence model, analyze and obtain visual features in character images, and associate the visual features with corresponding 3D spatial information to obtain associated information.

[0108] The feature extraction module is used to extract image features from human images based on the deep neural network in the artificial intelligence big model.

[0109] The three-dimensional reconstruction module is used to perform three-dimensional reconstruction of the person in the person image according to the associated information and image features to obtain a 3D virtual human model.

[0110] In some of these embodiments, the system further comprises: an identification module and an evaluation module.

[0111] The recognition module is used to recognize the fitness movements of the target user when the target user follows the movements and obtain the movement recognition results.

[0112] The evaluation module is used to evaluate the fitness movements of the target user based on the movement recognition results and the virtual human teaching animation.

[0113] In some embodiments, the evaluation module includes: a first position acquisition module, a second position acquisition module and a comparison module.

[0114] The first position acquisition module is used to obtain the first position information of the preset body skeleton key points of the target user according to the action recognition result.

[0115] The second position acquisition module is used to obtain second position information of preset body skeleton key points of the 3D virtual human model, and the 3D virtual human model is a 3D virtual human model in the virtual human teaching animation.

[0116] The comparison module is used to compare the first position information and the second position information to evaluate the fitness action of the target user.

[0117] In some embodiments, the animation generation module 42 includes: an information acquisition module, a matching module and a teaching animation generation module.

[0118] The information acquisition module is used to acquire the physical features of the target user from the character image and acquire the historical learning record of the target user when the character image is the character image of the target user.

[0119] The matching module is used to match the target fitness course according to the physical characteristics and / or historical learning records, and obtain the preset fitness skeleton animation corresponding to the target fitness course.

[0120] The teaching animation generation module is used to generate a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation of the target fitness course.

[0121] In some of the embodiments, the system further includes: a picture acquisition module.

[0122] The picture acquisition module is used to collect and save the scene pictures of the target user performing action following and learning. The scene pictures include the target user and the virtual human teaching animation.

[0123] Through the above system, the model building module 41 obtains the character image, generates a 3D virtual human model based on the character image through the artificial intelligence large model, and the animation generation module 42 generates a virtual human teaching animation based on the 3D virtual human model and the preset fitness skeleton animation. The display module 43 displays the virtual human teaching animation to the target user to instruct the target user to follow the movements according to the virtual human teaching animation, solving the problem that the recorded teaching video cannot fully and accurately convey the teaching content. The virtual human teaching animation can display the teaching process of the 3D virtual human model from various angles, and can convey the teaching content more comprehensively and accurately.

[0124] It should be noted that the above modules can be functional modules or program modules, and can be implemented by software or hardware. For modules implemented by hardware, the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0125] This embodiment further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0126] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0127] Optionally, in this embodiment, the processor may be configured to perform the following steps through a computer program:

[0128] S1, obtains a character image, and generates a 3D virtual human model based on the character image through an artificial intelligence large model.

[0129] S2, generating a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation.

[0130] S3, showing the virtual human teaching animation to the target user, so as to instruct the target user to follow the movements according to the virtual human teaching animation.

[0131] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0132] In one embodiment, Figure 5is a schematic diagram of the internal structure of an electronic device according to an embodiment of the present application, such as Figure 5 As shown, an electronic device is provided, which may be a server, and its internal structure diagram may be as shown in Figure 5 As shown. The electronic device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the electronic device is used to store data. The network interface of the electronic device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a fitness teaching method based on a virtual person is implemented.

[0133] Those skilled in the art will understand that Figure 5 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0134] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0135] Those skilled in the art should understand that the technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0136] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A fitness teaching method based on virtual human, characterized in that: The method comprises: Acquire a character image, and generate a 3D virtual human model based on the character image through an artificial intelligence model; Generate a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation; The virtual human teaching animation is displayed to the target user to instruct the target user to follow the movements of the virtual human teaching animation.

2. The method according to claim 1, characterized in that The displaying of the virtual human teaching animation to the target user comprises: Displaying the virtual human teaching screen through a projection screen or an AR device, wherein the virtual human teaching screen is a 3D screen; According to the angle adjustment instruction of the target user, the display angle of the 3D virtual human in the virtual human teaching screen is adjusted.

3. The method according to claim 1, characterized in that The step of generating a 3D virtual human model according to the character image by using the artificial intelligence large model comprises: Performing multimodal learning through the artificial intelligence big model, analyzing and acquiring visual features in the character image, and associating the visual features with corresponding 3D spatial information to obtain associated information; Extracting image features from the person image based on the deep neural network in the artificial intelligence big model; The person in the person image is three-dimensionally reconstructed according to the association information and the image features to obtain the 3D virtual human model.

4. The method according to claim 1, characterized in that The method further comprises: When the target user is performing action following, identifying the fitness action of the target user to obtain an action recognition result; The fitness movements of the target user are evaluated according to the movement recognition results and the virtual human teaching animation.

5. The method according to claim 4, characterized in that The evaluating the fitness action of the target user according to the action recognition result and the virtual human teaching animation comprises: Obtaining first position information of preset body skeleton key points of the target user according to the action recognition result; Acquire second position information of the preset body skeleton key point of the 3D virtual human model, wherein the 3D virtual human model is the 3D virtual human model in the virtual human teaching animation; The first location information and the second location information are compared to evaluate the fitness action of the target user.

6. The method according to claim 1, characterized in that The generating of the virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation comprises: In the case where the character image is a character image of the target user, obtaining the physical features of the target user from the character image, and obtaining the historical learning record of the target user; According to the physical features and / or the historical learning records, matching a target fitness course, and obtaining a preset fitness skeleton animation corresponding to the target fitness course; A virtual human teaching animation is generated according to the 3D virtual human model and the preset fitness skeleton animation of the target fitness course.

7. The method according to claim 1, characterized in that The method further comprises: The scene picture of the target user performing action following is collected and saved, wherein the scene picture includes the target user and the virtual human teaching animation.

8. A fitness teaching system based on virtual human, characterized in that: The system includes: a model building module, an animation generation module and a display module, wherein: The model building module is used to obtain a character image and generate a 3D virtual human model based on the character image through an artificial intelligence large model; The animation generation module is used to generate a virtual human teaching animation according to the 3D virtual human model and the preset fitness skeleton animation; The display module is used to display the virtual human teaching animation to the target user, so as to instruct the target user to follow the movements of the virtual human teaching animation.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the fitness teaching method based on a virtual person as described in any one of claims 1 to 7 is implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the fitness teaching method based on a virtual person as described in any one of claims 1 to 7 is implemented.

Images