CN114612600B - Virtual image generation method, device, electronic device and storage medium - Google Patents

Abstract

The disclosure provides a method for generating a virtual image, which relates to the technical field of artificial intelligence, and in particular to the technical fields of computer vision, virtual/augmented reality and metaverse. The specific implementation plan is as follows: convert multiple point data of the initial 3D virtual image into the frequency domain to obtain multiple frequency domain point data; render multiple frequency domain point data to generate the first 3D virtual image; determine the first 3D virtual image a perceptual characteristic of the avatar; and generating a second three-dimensional avatar based on a difference between the perceptual characteristic and a predetermined style characteristic. The present disclosure also provides a virtual image generating device, electronic equipment and a storage medium.

Description

Virtual image generation method, device, electronic device and storage medium

technical field

The present disclosure relates to the technical field of artificial intelligence, in particular to the technical fields of computer vision, virtual/augmented reality and metaverse, and can be applied in image processing scenarios. More specifically, the present disclosure provides a virtual image generation method, device, electronic device and storage medium.

Background technique

Avatars have a wide range of applications in scenarios such as metaverse, social networking, live streaming, or gaming. Avatars can be generated in a human-based manner.

Contents of the invention

The present disclosure provides a virtual image generation method, device, equipment and storage medium.

According to an aspect of the present disclosure, a method for generating virtual information is provided, the method comprising: converting multiple point data of an initial three-dimensional avatar into a frequency domain to obtain multiple frequency domain point data; adjusting the point data to obtain a plurality of adjusted point data; rendering the plurality of adjusted point data to generate a first three-dimensional virtual image; determining the perceptual characteristics of the first three-dimensional virtual image; and according to the The difference between the perceptual feature and the predetermined style feature is used to generate a second three-dimensional avatar.

According to another aspect of the present disclosure, there is provided an avatar generating device, the device comprising: a conversion module, configured to convert multiple point data of an initial three-dimensional avatar into a frequency domain to obtain multiple frequency domain point data; rendering A module for rendering the plurality of frequency domain point data to generate a first three-dimensional virtual image; a first determination module for determining the perceptual characteristics of the first three-dimensional virtual image; and a generating module for generating a first three-dimensional virtual image according to the A second three-dimensional avatar is generated based on the difference between the aforementioned perceptual features and the predetermined style features.

According to another aspect of the present disclosure, an electronic device is provided, including: at least one processor; and a memory communicated with the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by At least one processor executes, so that the at least one processor can execute the method provided according to the present disclosure.

According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the method provided according to the present disclosure.

According to another aspect of the present disclosure, there is provided a computer program product comprising a computer program which, when executed by a processor, implements the method provided according to the present disclosure.

It should be understood that what is described in this section is not intended to identify key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will be readily understood through the following description.

Description of drawings

The accompanying drawings are used to better understand the present solution, and do not constitute a limitation to the present disclosure. in:

Fig. 1 is a schematic diagram of an exemplary system architecture that can be applied to an avatar generating method and device according to an embodiment of the present disclosure;

FIG. 2 is a flowchart of a method for generating an avatar according to an embodiment of the present disclosure;

FIG. 3 is a flowchart of a method for generating an avatar according to an embodiment of the present disclosure;

FIG. 4 is a flowchart of a method for generating an avatar according to an embodiment of the present disclosure;

5 is a block diagram of an avatar generating device according to an embodiment of the present disclosure; and

FIG. 6 is a block diagram of an electronic device to which an avatar generating method can be applied according to one embodiment of the present disclosure.

Detailed ways

The exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and they should be regarded as exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

An avatar may include a virtual body. An avatar can be designed, generated and optimized manually, which requires a high time cost. Moreover, the avatar style generated based on artificial methods is relatively single.

Fig. 1 is a schematic diagram of an exemplary system architecture in which the method and device for generating an avatar can be applied according to an embodiment of the present disclosure. It should be noted that Figure 1 is only an example of the system architecture to which the embodiments of the present disclosure can be applied, to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other device, system, environment or scenario.

As shown in FIG. 1 , a system architecture 100 according to this embodiment may include terminal devices 101 , 102 , 103 , a network 104 and a server 105 . The network 104 is used as a medium for providing communication links between the terminal devices 101 , 102 , 103 and the server 105 . Network 104 may include various connection types, such as wired and/or wireless communication links, among others.

A user may interact with a server 105 over a network 104 using a terminal device 101, 102, 103 to receive or send messages or the like. The terminal devices 101, 102, 103 may be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers and the like.

The server 105 may be a server that provides various services, such as a background management server that provides support for websites browsed by users using the terminal devices 101, 102, 103 (just an example). The background management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

It should be noted that, generally, the method for generating an avatar provided by the embodiment of the present disclosure can be executed by the server 105 . Correspondingly, the virtual image generation device provided by the embodiment of the present disclosure can generally be set in the server 105. The avatar generating method provided by the embodiments of the present disclosure may also be executed by a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and/or the server 105. Correspondingly, the avatar generating apparatus provided by the embodiments of the present disclosure may also be set in a server or a server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and/or the server 105.

Fig. 2 is a flowchart of a method for generating an avatar according to an embodiment of the present disclosure.

As shown in FIG. 2 , the method 200 may include operation S210 to operation S240.

In operation S210, a plurality of point data of the initial three-dimensional avatar is converted into a frequency domain to obtain a plurality of frequency domain point data.

For example, the initial three-dimensional virtual image may be a preset three-dimensional virtual image.

For example, a Fourier transform may be performed on multiple point data on the initial three-dimensional avatar to transform the multiple point data into the frequency domain.

In operation S220, a plurality of frequency-domain point data are rendered to generate a first three-dimensional virtual image.

For example, multiple frequency domain point data can be rendered using various renderers. In one example, multiple frequency domain point data can be rendered using the Pytorch3D renderer.

In operation S230, perceptual characteristics of the first three-dimensional avatar are determined.

For example, the perceptual features of the first 3D virtual learning can be determined according to various feature extraction models.

In operation S240, a second 3D avatar is generated according to the difference between the perceptual feature and the predetermined style feature.

For example, various feature extraction models can be used to perform feature feature extraction on the acquired style description information to obtain predetermined style features.

For example, various loss functions can be used to determine the difference value between perceptual features and predetermined stylistic features. In one example, the difference value between the perceptual feature and the predetermined style feature can be determined according to the L2 loss function. If the difference value satisfies the predetermined condition, the above-mentioned first three-dimensional virtual image can be used as the second three-dimensional virtual image. If the difference value does not meet the predetermined condition, the above-mentioned first three-dimensional avatar may be adjusted until the difference between the perceptual feature and the predetermined style feature meets the predetermined condition. For example, the predetermined condition may be that the difference value is smaller than a predetermined threshold.

Through the embodiments of the present disclosure, a three-dimensional virtual image matching predetermined style characteristics can be generated.

In some embodiments, the first three-dimensional virtual image can be processed by using the contrastive image-text pre-training model to obtain the perceptual features of the first three-dimensional virtual image.

For example, the Contrastive Language-Image Pre-training (CLIP) model can extract features of text as well as features of images. Contrastive Text Pre-Training Model is an open-source general-purpose model that connects text and images. The task of the contrastive image-text pre-training model is to identify various visual information in the image and associate this information with one of the massive images.

In an example, a screenshot operation may be performed on the display screen displaying the first three-dimensional avatar to obtain a screenshot image. The screenshot image is processed by using the contrastive image-text pre-training model to obtain perceptual features.

In some embodiments, the predetermined style features are determined according to the style description information by using the pre-trained model of contrasting pictures and texts.

For example, it is possible to obtain a text input by the target object and use the text as a style description information. Next, the style description information can be processed using the above-mentioned contrastive image-text pre-training model to determine predetermined style features. In one example, the style description information may be, for example, text including keywords such as "cute" and "cool". Comparing the image-text pre-training model can efficiently determine whether the image and text match. In addition, the predetermined style features and perceptual features can be determined by the same pre-training model of contrasting images and texts. After adjusting the differences between the two, the two can be more matched, so as to generate a 3D virtual image that is more in line with the style description information.

In some embodiments, the plurality of point data of the initial three-dimensional avatar is converted to the frequency domain, so that various 3D tools can be used for processing based on the plurality of frequency domain point data. For example, the 3D tool can be a Unity 3D tool.

In some embodiments, the above-mentioned predetermined condition may be that the difference values converge.

Fig. 3 is a flowchart of a method for generating an avatar according to another embodiment of the present disclosure.

As shown in FIG. 3 , the method 300 may include operation S310 to operation S330, and operation S341 to operation S344.

In operation S310, a plurality of point data of the initial three-dimensional avatar is converted into a frequency domain to obtain a plurality of frequency domain point data.

For example, operation S310 is the same as or similar to operation S210 described above, and will not be repeated in this disclosure.

In operation S320, a plurality of frequency-domain point data are rendered to generate a first three-dimensional virtual image.

For example, operation S320 is the same as or similar to operation S220 described above, and will not be repeated in this disclosure.

In operation S330, perceptual characteristics of the first three-dimensional avatar are determined.

For example, the perceptual characteristics of the first three-dimensional virtual image can be determined by using the above-mentioned contrastive image-text pre-training model.

In operation S341, a difference value between the perceptual feature and the predetermined style feature is determined.

For example, the predetermined style feature is determined according to the style description information by using the above-mentioned contrastive image-text pre-training model.

For example, the L2 loss function can be used to determine the difference value between the perceptual feature and the predetermined style feature. The L2 loss function is also called the Least Square Error (LSE) loss function.

In operation S342, it is determined whether the difference values converge.

In the embodiment of the present disclosure, in a case where it is determined that the difference values converge, operation S343 is performed.

For example, after it is determined that the nth difference value is less than or equal to the predetermined difference threshold, if it is determined that the i difference values after the nth difference value are all less than or equal to the predetermined difference threshold, it can be determined that the difference values converge. In one example, n is an integer greater than or equal to 1, and i is an integer greater than or equal to 1. For example, i is a preset value, i=1.

In the embodiment of the present disclosure, if it is determined that the difference value does not converge, perform operation S344, and then return to operation S320.

For example, after determining that the m-th difference value is less than or equal to a predetermined difference threshold, if any one of the j difference values after the m-th difference value is determined to be greater than the preset difference threshold, it can be determined that the difference value does not converge, Operation S344 may be performed. After performing operation S344, it may return to operation S320. In one example, m is an integer greater than or equal to 1, and j is an integer greater than or equal to 1. For example, j is a preset value, j=1.

In operation S343, the current first three-dimensional virtual image is used as the second three-dimensional virtual image

For example, as described above, after it is determined that the difference values converge, the first 3D avatar Vir_n corresponding to the nth difference value may be used as the second 3D avatar.

In operation S344, a plurality of frequency domain point data are adjusted.

For example, as described above, after it is determined that the difference value does not converge, multiple frequency domain point data corresponding to the m+jth difference value may be adjusted to obtain multiple adjusted frequency domain point data. Based on the multiple adjusted frequency domain point data, return to operation S320, render the multiple adjusted frequency domain point data, and generate the m+j+1th first three-dimensional virtual image. Then perform subsequent operations.

For example, point data includes point coordinate data and color data. In one example, the frequency domain point data includes frequency domain point coordinate data and frequency domain color data.

Through the embodiments of the present disclosure, when the difference value does not converge, the frequency domain point data is adjusted until the difference value converges, so that the perceptual characteristics of the second 3D avatar match the predetermined style characteristics, thereby improving user experience.

In some embodiments, unlike method 300, the difference value may be compared to a preset difference threshold to determine whether the difference value has converged.

For example, in a case where the nth difference value is less than or equal to a preset threshold, it is determined that the difference values converge.

For another example, when the nth difference value is greater than a preset difference threshold, it is determined that the difference value does not converge.

Fig. 4 is a flowchart of a method for generating an avatar according to another embodiment of the present disclosure.

As shown in Fig. 4, the method 444 can adjust multiple frequency-domain point data, which will be described in detail below in conjunction with operation S4441-operation S4442.

In operation S4441, for each of the plurality of frequency-domain point data, a point normal of each frequency-domain point data is determined.

For example, the Unity 3D tool can be used to determine a mesh (Mesh) model Model_Mesh_k based on multiple frequency domain point data. The grid model Model_Mesh_k contains multiple triangular plane slice models. In one example, a triangular planar patch submodel may correspond to a frequency-domain point data. In one example, a 3D avatar can be obtained by rendering according to the mesh model Model_Mesh_k.

In operation S4442, each frequency-domain point data is adjusted in a direction in which the point normal extends.

For example, as mentioned above, the frequency-domain point data includes frequency-domain point coordinate data and frequency-domain point color data. The value of the frequency-domain point coordinate data can be adjusted along the direction in which the normal line of the point extends.

For example, after adjusting the value of point coordinate data along the direction of point normal extension, an adjusted mesh model Model_Mesh_k+1 can be obtained. k is an integer greater than or equal to 1.

In an example, the mesh model Model_Mesh_k+1 is rendered by using a renderer to obtain the first 3D avatar after the k+1th adjustment. Adjusting along the direction of point normal extension can ensure that each frequency domain point moves within a certain range, so that the adjusted frequency domain point data distribution is more uniform.

In some embodiments, adjusting the plurality of frequency-domain point data may include: adjusting the value of each frequency-domain color data.

In some embodiments, the data structure of the grid model may be a graph structure. Accordingly, the mesh model may include multiple points, multiple edges, and multiple faces.

For example, the data structure of the grid model can be a directed graph structure. For another example, the data structure of the grid model can be an undirected graph structure.

In some embodiments, point normals may be vertex normals.

For example, the face domain normals of the vertices of the triangular planar slice can be weighted average to obtain the vertex normals.

In some embodiments, different from the method 400, adjusting the multiple frequency domain point data includes: determining the surface point normal according to the multiple frequency domain point data; adjusting the multiple frequency domain point data along the extending direction of the surface point normal .

For example, a face may be defined from at least one triangular planar piece. Surface normals can represent vertices within a surface rather than vertices of a mesh model. The relationship between surface normals and mesh vertices is a many-to-one relationship. For example, for a corner point in a cube mesh model, the corner point has three perpendicular adjacent faces. The face point normal can be determined from these three perpendicular adjacent faces.

FIG. 5 is a block diagram of an avatar generating device according to an embodiment of the present disclosure.

As shown in FIG. 5 , the apparatus 500 may include a conversion module 510, a rendering module 520, a first determination module 530 and a generation module 540.

The conversion module 510 is configured to convert multiple point data of the initial 3D avatar into frequency domain to obtain multiple frequency domain point data. In an example, the conversion module 510 may be used to perform, for example, operation S210 in FIG. 2 .

The rendering module 520 is configured to render the plurality of frequency-domain point data to generate a first three-dimensional virtual image. In an example, the rendering module 520 may be used to perform, for example, operation S220 in FIG. 2 .

The first determination module 530 is configured to determine the perceptual features of the first 3D avatar. In an example, the first determination module 530 may be used to perform, for example, operation S230 in FIG. 2 .

A generating module 540, configured to generate a second 3D avatar according to the difference between the perceptual feature and a predetermined style feature. In an example, the generating module 540 may be used to perform, for example, operation S240 in FIG. 2 .

In some embodiments, the generation module includes: a first determination submodule, configured to determine a difference value between the perceptual feature and a predetermined style feature; a second determination submodule, used to determine whether the difference value converges ; Obtaining a submodule, used to use the current first 3D avatar as the second 3D avatar when it is determined that the difference value is convergent; and an adjustment submodule, used to determine that the difference value does not converge In some cases, the multiple frequency domain point data are adjusted, and the operation of rendering the multiple frequency domain point data is returned.

In some embodiments, the point data includes point coordinate data and color data.

In some embodiments, the adjustment submodule includes: a determining unit, configured to, for each frequency domain point data in the plurality of frequency domain point data, determine a point normal of each frequency domain point data; and an adjustment unit, configured to adjust each of the frequency-domain point data along the direction in which the normal line of the point extends.

In some embodiments, the first determination module includes: processing the first 3D avatar with a contrastive image-text pre-training model to obtain the perceptual features of the first 3D avatar.

In some embodiments, the device 500 further includes: a second determining module, configured to determine the predetermined style features according to the style description information by using the contrastive image-text pre-training model.

In the technical solution of this disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved are all in compliance with relevant laws and regulations, and do not violate public order and good customs.

According to the embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

For example, the present disclosure also provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by At least one processor executes, so that the at least one processor can execute the method provided according to the present disclosure.

For example, the present disclosure also provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the method according to the present disclosure.

For example, the present disclosure also provides a computer program product comprising a computer program that, when executed by a processor, implements the method provided according to the present disclosure. The following will be described in detail in conjunction with Figure 6.

Figure 6 shows a schematic block diagram of an example electronic device 600 that may be used to implement embodiments of the present disclosure. Electronic device is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. Electronic devices may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are by way of example only, and are not intended to limit implementations of the disclosure described and/or claimed herein.

As shown in FIG. 6, the device 600 includes a computing unit 601 that can execute according to a computer program stored in a read-only memory (ROM) 602 or loaded from a storage unit 608 into a random-access memory (RAM) 603. Various appropriate actions and treatments. In the RAM 603, various programs and data necessary for the operation of the device 600 can also be stored. The computing unit 601, ROM 602, and RAM 603 are connected to each other through a bus 604. An input/output (I/O) interface 605 is also connected to bus 604 .

Multiple components in the device 600 are connected to the I/O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a magnetic disk, an optical disk, etc. ; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, and the like. The communication unit 609 allows the device 600 to exchange information/data with other devices over a computer network such as the Internet and/or various telecommunication networks.

Computing unit 601 may be various general-purpose and/or special-purpose processing components having processing and computing capabilities. Some examples of computing units 601 include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, digital signal processing processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 executes various methods and processes described above, such as a method for generating an avatar. For example, in some embodiments, the avatar generating method may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device 600 via the ROM 602 and/or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the avatar generating method described above can be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to execute the avatar generating method in any other suitable manner (for example, by means of firmware).

Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips Implemented in a system of systems (SOC), complex programmable logic device (CPLD), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include being implemented in one or more computer programs executable and/or interpreted on a programmable system including at least one programmable processor, the programmable processor Can be special-purpose or general-purpose programmable processor, can receive data and instruction from storage system, at least one input device, and at least one output device, and transmit data and instruction to this storage system, this at least one input device, and this at least one output device an output device.

Program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special purpose computer, or other programmable data processing devices, so that the program codes, when executed by the processor or controller, make the functions/functions specified in the flow diagrams and/or block diagrams Action is implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media would include one or more wire-based electrical connections, portable computer disks, hard disks, Random Access Memory (RAM), Read Only Memory (ROM), Erasable Programmable Read Only Memory (EPROM or flash memory), optical fiber, compact disk read only memory (CD-ROM), optical storage, magnetic storage, or any suitable combination of the foregoing.

To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user. ); and a keyboard and pointing device (eg, a mouse or a trackball) through which a user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and can be in any form (including Acoustic input, speech input, or, tactile input) to receive input from the user.

The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., as a a user computer having a graphical user interface or web browser through which a user can interact with embodiments of the systems and techniques described herein), or including such backend components, middleware components, Or any combination of front-end components in a computing system. The components of the system can be interconnected by any form or medium of digital data communication, eg, a communication network. Examples of communication networks include: Local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.

A computer system may include clients and servers. Clients and servers are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by computer programs running on the respective computers and having a client-server relationship to each other.

It should be understood that steps may be reordered, added or deleted using the various forms of flow shown above. For example, each step described in the present disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired result of the technical solution disclosed in the present disclosure can be achieved, no limitation is imposed herein.

The specific implementation manners described above do not limit the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions may be made depending on design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present disclosure shall be included within the protection scope of the present disclosure.

Claims (12)

1. A virtual image generation method, comprising: converting multiple point data of the initial three-dimensional avatar into the frequency domain to obtain multiple frequency domain point data; Rendering the plurality of frequency-domain point data to generate a first three-dimensional virtual image; determining perceptual characteristics of the first 3D avatar; and generating a second three-dimensional avatar based on the difference between the perceptual feature and a predetermined style feature, Wherein, the generating the second three-dimensional avatar according to the difference between the perceptual feature and the predetermined style feature includes: determining a difference value between said perceptual characteristic and a predetermined stylistic characteristic; determining whether the difference values converge; When it is determined that the difference value does not converge, adjust the multiple frequency domain point data, and return the operation of rendering the multiple frequency domain point data; Wherein, said adjusting said multiple frequency domain point data includes: For each frequency domain point data in the plurality of frequency domain point data, determining the point normal of each of the frequency domain point data; and Adjusting each of the frequency domain point data along a direction in which a normal line of the point extends. 2. The method according to claim 1, wherein said generating a second three-dimensional avatar based on the difference between said perceptual features and predetermined style features further comprises: If it is determined that the difference values converge, the current first three-dimensional virtual image is used as the second three-dimensional virtual image. 3. The method according to claim 1 or 2, wherein the point data comprises point coordinate data and color data. 4. The method according to claim 1, wherein said determining the perceptual characteristics of said first three-dimensional avatar comprises: The first three-dimensional virtual image is processed by using the pre-training model of contrasting images to obtain the perceptual features of the first three-dimensional virtual image. 5. The method of claim 1, further comprising: The predetermined style feature is determined according to the style description information by using a pre-trained model for comparing images and texts. 6. A virtual image generating device, comprising: A conversion module is used to convert multiple point data of the initial three-dimensional virtual image into the frequency domain to obtain multiple frequency domain point data; A rendering module, configured to render the plurality of frequency-domain point data to generate a first three-dimensional virtual image; A first determining module, configured to determine the perceptual characteristics of the first three-dimensional avatar; and A generating module, configured to generate a second three-dimensional avatar according to the difference between the perceptual feature and the predetermined style feature; Wherein, the generating module includes: A first determining submodule, configured to determine a difference value between the perceptual feature and a predetermined style feature; The second determining submodule is used to determine whether the difference value is converged; An adjustment submodule, configured to adjust the plurality of frequency-domain point data when it is determined that the difference value does not converge, and return the operation of rendering the plurality of frequency-domain point data; The adjustment sub-module includes: A determining unit, configured to, for each frequency domain point data in the plurality of frequency domain point data, determine a point normal of each frequency domain point data; and An adjustment unit, configured to adjust each of the frequency-domain point data along a direction in which the normal line of the point extends. 7. The apparatus according to claim 6, wherein the generating module comprises: The obtaining submodule is configured to use the current first 3D avatar as the second 3D avatar when it is determined that the difference values converge. 8. The apparatus according to claim 6 or 7, wherein the point data comprises point coordinate data and color data. 9. The device according to claim 6, wherein the first determining module is further configured to: The first three-dimensional virtual image is processed by using the pre-training model of contrasting images to obtain the perceptual features of the first three-dimensional virtual image. 10. The apparatus of claim 6, further comprising: The second determination module is configured to determine the predetermined style feature according to the style description information by using the pre-training model of contrasting pictures and texts. 11. An electronic device comprising: at least one processor; and a memory communicatively coupled to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions are executed by the at least one processor, so that the at least one processor can perform any one of claims 1 to 5 Methods. 12. A non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the method according to any one of claims 1-5.

Images