Model rendering method, apparatus, and device

Abstract

The present disclosure provides a model rendering method, device and equipment, relates to the technical field of artificial intelligence, specifically to the technical field of augmented reality, deep learning and computer vision, and can be applied to the scene of rendering photorealistic digital people. A specific embodiment of the method comprises: learning a white model of a first precision in a first engine; learning multi-angle map data of a second precision in a second engine, wherein the second precision is higher than the first precision, and mapping the multi-angle map data to the white model to obtain a rendering model. The embodiment reduces the requirement of rendering on hardware computing power and can realize real-time end-side rendering.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, specifically augmented reality, deep learning, and computer vision technology, and can be applied to rendering hyper-realistic digital human scenarios. Background Technology

[0002] With the rapid iteration of artificial intelligence and augmented reality technologies, digital humans are also flourishing. However, hyper-realistic digital humans are limited by the current hardware's computing power, making real-time edge rendering impossible. Cartoon digital humans, on the other hand, cannot showcase the richness of hyper-realistic digital humans. Therefore, the current method for rendering hyper-realistic digital humans typically involves transmitting video streams to a server for rendering via high-speed, low-latency broadband. Summary of the Invention

[0003] This disclosure provides a model rendering method, apparatus, device, storage medium, and program product.

[0004] In a first aspect, embodiments of this disclosure propose a model rendering method, comprising: learning a white model of a first precision in a first engine; learning multi-angle texture data of a second precision in a second engine, wherein the second precision is higher than the first precision; mapping the multi-angle texture data onto the white model to obtain a rendered model.

[0005] Secondly, embodiments of this disclosure propose a model rendering apparatus, comprising: a first learning module configured to learn a white model of a first precision in a first engine; a second learning module configured to learn multi-angle texture data of a second precision in a second engine, wherein the second precision is higher than the first precision; and a rendering module configured to map the multi-angle texture data onto the white model to obtain a rendered model.

[0006] Thirdly, embodiments of this disclosure provide an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a method as described in any implementation of the first aspect.

[0007] Fourthly, embodiments of this disclosure provide a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform a method as described in any implementation of the first aspect.

[0008] Fifthly, embodiments of this disclosure provide a computer program product, including a computer program that, when executed by a processor, implements the method as described in any of the implementations in the first aspect.

[0009] The model rendering method provided in this disclosure reduces the hardware computing power requirements for rendering, enables real-time edge rendering, and can be applied to rendering hyper-realistic digital human scenarios.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0011] Other features, objects, and advantages of this disclosure will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings. The drawings are provided for a better understanding of the invention and are not intended to limit the scope of this disclosure. Wherein:

[0012] Figure 1 This is a flowchart of an embodiment of the model rendering method according to the present disclosure;

[0013] Figure 2 This is a flowchart of yet another embodiment of the model rendering method according to this disclosure;

[0014] Figure 3 This is a flowchart of another embodiment of the model rendering method according to the present disclosure;

[0015] Figure 4 This is a block diagram of rendering a hyper-realistic digital human using the model rendering method engine of this embodiment;

[0016] Figure 5 This is another block diagram of rendering hyper-realistic digital humans using the model rendering method engine of the present disclosure embodiments;

[0017] Figure 6 This is a schematic diagram of the structure of an embodiment of the model rendering apparatus according to the present disclosure;

[0018] Figure 7 This is a block diagram of an electronic device used to implement the model rendering method of the embodiments of this disclosure. Detailed Implementation

[0019] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other. This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] Figure 1 A flow 100 of an embodiment of a model rendering method according to the present disclosure is shown. The model rendering method includes the following steps:

[0022] Step 101: Learn the first-precision white model in the first engine.

[0023] In this embodiment, the execution body of the model rendering method can learn the first-precision white model in the first engine.

[0024] The white model can be an unrendered 3D model, such as an unrendered hyper-realistic digital human. The first engine can be a rendering engine used to render the model. For example, the first engine could be a self-developed rendering engine with lower precision than common rendering engines. Therefore, the first engine learns a low-precision white model, which can use standard PBR (Physically Based Rendering) materials.

[0025] Step 102: Learn the second-precision multi-angle texture data in the second engine.

[0026] In this embodiment, the aforementioned execution entity can learn multi-angle texture data of a second precision from the second engine. The second precision is higher than the first precision.

[0027] The multi-angle texture data can be obtained by extracting texture data from different angles from a reference rendering model in the second engine. The reference rendering model can be, for example, a rendered hyper-realistic digitized human in the second engine. Typically, a large amount of multi-angle texture data can be obtained by extracting textures from different angles from a large number of rendered hyper-realistic digitized humans. The second engine can also be a rendering engine used to render the model. For example, the second engine could be UE5 (Unreal Engine 5), which has higher precision than the first engine. Therefore, the second engine learns high-precision multi-angle texture data, which can be used to extend the skin material using PBR.

[0028] Step 103: Map the multi-angle texture data onto the white model to obtain the rendered model.

[0029] In this embodiment, the aforementioned execution entity can map multi-angle texture data onto the white model to obtain the rendered model.

[0030] Typically, multi-angle texture data can be fitted to the white model on the first engine based on a large amount of multi-angle texture data. In the first engine, the fitted multi-angle texture data is used to render the white model, thus completing the rendering of the white model on the edge engine and obtaining the rendered model.

[0031] The model rendering method provided in this disclosure reduces the hardware computing power requirements for rendering, enables real-time edge rendering, and can be applied to rendering hyper-realistic digital human scenarios.

[0032] Continue to refer to Figure 2 This illustrates a flow 200 of yet another embodiment of the model rendering method according to the present disclosure. The model rendering method includes the following steps:

[0033] Step 201: Learn the first-precision white model in the first engine.

[0034] In this embodiment, the execution body of the model rendering method can learn the first-precision white model in the first engine.

[0035] The white model can be an unrendered 3D model, such as an unrendered hyper-realistic digital human. The first engine can be a rendering engine used to render the model. For example, the first engine could be a self-developed rendering engine with lower precision than common rendering engines. Therefore, the first engine learns a low-precision white model, which can use standard PBR materials.

[0036] Step 202: Extract reference rendering models from different angles in the second engine to obtain reference multi-angle texture data.

[0037] In this embodiment, the aforementioned execution entity can extract reference rendering models from different angles in the second engine to obtain reference multi-angle texture data.

[0038] The second engine contains a large number of pre-rendered reference models. These reference models can be, for example, hyper-realistic digitized humans rendered in the second engine. Typically, extracting data from different angles from these pre-rendered hyper-realistic digitized humans yields a large amount of reference multi-angle texture data. Because the second engine has higher precision than the first engine, it learns high-precision reference multi-angle texture data.

[0039] Step 203: The reference multi-angle texture data and the white model are fused and trained to obtain multi-angle texture data.

[0040] In this embodiment, the aforementioned execution entity can fuse and train the reference multi-angle texture data and the white model to obtain multi-angle texture data.

[0041] Typically, by extracting different angles from a large number of reference rendering models in the second engine and fusing them with the white model in the first engine, a large amount of multi-angle texture data can be obtained, which can be used to fit the rendering data of the white model in the first engine.

[0042] Step 204: Fit the white model and multi-angle texture data to generate rendering data that couples the textures with the white model in the first engine.

[0043] In this embodiment, the aforementioned execution entity can fit the white model and multi-angle texture data to generate rendering data in the first engine that couples the textures with the white model.

[0044] Typically, the white model from the first engine and the multi-angle texture data from the second engine are trained together for fitting. Both engines use standard PBR materials, producing rendering data that couples the textures from the first engine with the white model.

[0045] Step 205: Input the rendering data into the first engine to render the white model and obtain the rendered model.

[0046] In this embodiment, the aforementioned execution entity can input rendering data into the first engine to render the white model and obtain the rendered model.

[0047] Typically, the rendering data is placed in the first engine. The first engine can perform rendering using the rendering data and the white model, thus completing the rendering of the white model on the end engine and obtaining the rendered model.

[0048] from Figure 2 It can be seen from this that, with Figure 1 Compared to the corresponding embodiments, the model rendering method flow 200 in this embodiment emphasizes the texture extraction step and the rendering step. Therefore, the solution described in this embodiment not only improves the matching degree between multi-angle texture data and the white model, but also enhances the model rendering effect, and can be applied to rendering hyper-realistic digital human scenarios.

[0049] Further reference Figure 3 This illustrates a flow 300 of another embodiment of the model rendering method according to the present disclosure. The model rendering method includes the following steps:

[0050] Step 301: Learn the first-precision white model in the first engine.

[0051] In this embodiment, the execution body of the model rendering method can learn the first-precision white model in the first engine.

[0052] The white model can be an unrendered 3D model, such as an unrendered hyper-realistic digital human. The first engine can be a rendering engine used to render the model. For example, the first engine could be a self-developed rendering engine with lower precision than common rendering engines. Therefore, the first engine learns a low-precision white model, which can use standard PBR materials.

[0053] Step 302: Extract reference rendering models from different angles in the second engine to obtain reference multi-angle texture data.

[0054] In this embodiment, the aforementioned execution entity can extract reference rendering models from different angles in the second engine to obtain reference multi-angle texture data.

[0055] The second engine contains a large number of pre-rendered reference models. These reference models can be, for example, hyper-realistic digitized humans rendered in the second engine. Typically, extracting data from different angles from these pre-rendered hyper-realistic digitized humans yields a large amount of reference multi-angle texture data. Because the second engine has higher precision than the first engine, it learns high-precision reference multi-angle texture data.

[0056] Step 303: Determine at least one of the target lighting, target viewpoint, and target pose texture space of the white model.

[0057] In this embodiment, the aforementioned execution entity can determine at least one of the target lighting, target viewpoint, and target pose texture space of the white model.

[0058] In this context, target lighting refers to the lighting of the model to be rendered from the white model. Target viewpoint refers to the viewpoint of the model to be rendered from the white model. Target pose texture space refers to the pose texture space of the model to be rendered from the white model. Texture space is chosen over image space because the hyper-realistic digital human has multiple materials such as skin, hair, and clothing, making segmentation impossible in image space. Choosing texture space allows for precise control over facial skin.

[0059] Step 304: Set the reference multi-angle texture data to be the same as at least one of the target lighting, target viewpoint and target pose texture space to generate multi-angle texture data.

[0060] In this embodiment, the aforementioned execution entity can set the reference multi-angle texture data to be the same as at least one of the target lighting, target viewpoint, and target pose texture space to generate multi-angle texture data.

[0061] Typically, by controlling the same lighting, viewpoint, and pose texture space, multiple sets of multi-angle texture data that better match the model to be rendered can be generated.

[0062] Step 305: Fit the white model and multi-angle texture data to generate rendering data that couples the textures with the white model in the first engine.

[0063] In this embodiment, the aforementioned execution entity can fit the white model and multi-angle texture data to generate rendering data in the first engine that couples the textures with the white model.

[0064] Typically, the white model from the first engine and the multi-angle texture data from the second engine are trained together for fitting. Both engines use standard PBR materials, producing rendering data that couples the textures from the first engine with the white model.

[0065] Step 306: Input the rendering data into the facial wireless mesh network, and sequentially perform facial expression driving, bone driving and unfolding shading on the face of the white model to obtain the facial rendering model.

[0066] In this embodiment, the aforementioned execution entity can input rendering data into a facial wireless mesh network, and sequentially perform facial expression driving, bone driving, and unfolding shading on the face of the white model to obtain a facial rendering model.

[0067] Step 307: Input the facial rendering model into the overall wireless mesh network, and sequentially perform facial expression driving, bone driving, and physically based shading and no-light shading on the facial rendering model to obtain the rendering model.

[0068] In this embodiment, the aforementioned execution entity can input the facial rendering model into the overall wireless mesh network, and sequentially perform facial expression driving, skeletal driving, and physically based shading and no-light shading on the facial rendering model to obtain the rendering model.

[0069] Typically, the first engine can include a facial wireless mesh network and an overall wireless mesh network. Rendering data can be submitted to the facial mesh, where facial expression-driven (BS), skeletal-driven, and PBR-based shading are performed sequentially to obtain the facial rendering model. The facial rendering model can be transferred to the overall mesh via Pix2Pix, where BS-driven, skeletal-driven, PBR, and unlit shading are performed sequentially to obtain the rendered model. Performing facial rendering first, followed by overall rendering, improves the detail of the model rendering. Pix2Pix is ​​a classic image transfer method that can be used in scenarios where shape is defined and image quality is enhanced.

[0070] from Figure 3 It can be seen from this that, with Figure 2Compared to the corresponding embodiments, the model rendering method in this embodiment emphasizes the texture extraction and rendering steps in process flow 300. Therefore, the solution described in this embodiment not only improves the matching degree between multi-angle texture data and the desired rendered model, but also enhances the detail of the model rendering, making it applicable to hyper-realistic digital human rendering scenarios.

[0071] For ease of understanding, Figure 4 A block diagram is shown illustrating the rendering of a hyper-realistic digital human using a model rendering method engine according to an embodiment of the present disclosure. Figure 4 As shown, the first engine can include the entire mesh. Rendering data can be submitted to the entire mesh, and then processed sequentially using BS-driven, skeleton-driven, and PBR shading techniques to display the resulting rendered model on screen.

[0072] For ease of understanding, Figure 5 Another block diagram is shown illustrating the rendering of a hyper-realistic digital human using a model rendering method engine according to embodiments of the present disclosure. For example... Figure 5 As shown, the first engine can include a facial mesh and an overall mesh. Rendering data can be submitted to the facial mesh, where it undergoes BS-driven, skeletal-driven, and PBR shading processes in sequence to obtain the facial rendering model. The facial rendering model can be transferred to the overall mesh via Pix2Pix, where it undergoes BS-driven, skeletal-driven, and PBR shading processes, along with unlit shading, to display the resulting rendering model on screen.

[0073] Further reference Figure 6 As an implementation of the methods shown in the above figures, this disclosure provides an embodiment of a model rendering apparatus, which is similar to... Figure 1 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0074] like Figure 6 As shown, the model rendering device 400 of this embodiment may include: a first learning module 601, a second learning module 602, and a rendering module 603. The first learning module 601 is configured to learn a white model of a first precision from a first engine; the second learning module 602 is configured to learn multi-angle texture data of a second precision from a second engine, wherein the second precision is higher than the first precision; and the rendering module 603 is configured to map the multi-angle texture data onto the white model to obtain a rendered model.

[0075] In this embodiment, the specific processing of the first learning module 601, the second learning module 602, and the rendering module 603 in the model rendering device 600, and the resulting technical effects, can be found in reference to [reference needed]. Figure 1 The relevant descriptions of steps 101-103 in the corresponding embodiments will not be repeated here.

[0076] In some optional implementations of this embodiment, the second learning module 602 includes: an extraction submodule, configured to extract reference rendering models in the second engine from different angles to obtain reference multi-angle texture data; and a fusion submodule, configured to fuse the reference multi-angle texture data and the white model for training to obtain multi-angle texture data.

[0077] In some optional implementations of this embodiment, the fusion submodule is further configured to: determine the target lighting, target viewpoint, and target pose texture space of the white model; set the reference multi-angle texture data to be the same as the target lighting, target viewpoint, and target pose texture space, and generate multi-angle texture data.

[0078] In some optional implementations of this embodiment, the rendering module 603 includes: a fitting submodule configured to fit the white model and multi-angle texture data to generate rendering data in the first engine that couples the texture with the white model; and a rendering submodule configured to input the rendering data into the first engine to render the white model and obtain a rendering model.

[0079] In some optional implementations of this embodiment, the rendering submodule is further configured to: input rendering data into the facial wireless mesh network, sequentially perform facial expression driving, bone driving, and unfolded shading on the face of the white model to obtain a facial rendering model; input the facial rendering model into the overall wireless mesh network, sequentially perform facial expression driving, bone driving, and physically based shading and no-light shading on the facial rendering model to obtain a rendering model.

[0080] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0081] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0082] Figure 7 A schematic block diagram of an example electronic device 700 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0083] like Figure 7 As shown, device 700 includes a computing unit 701, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 702 or a computer program loaded from storage unit 708 into random access memory (RAM) 703. RAM 703 may also store various programs and data required for the operation of device 700. The computing unit 701, ROM 702, and RAM 703 are interconnected via bus 704. Input / output (I / O) interface 705 is also connected to bus 704.

[0084] Multiple components in device 700 are connected to I / O interface 705, including: input unit 706, such as keyboard, mouse, etc.; output unit 707, such as various types of monitors, speakers, etc.; storage unit 708, such as disk, optical disk, etc.; and communication unit 709, such as network card, modem, wireless transceiver, etc. Communication unit 709 allows device 700 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0085] The computing unit 701 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 701 performs the various methods and processes described above, such as the model rendering method. For example, in some embodiments, the model rendering method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 708. In some embodiments, part or all of the computer program may be loaded and / or installed on device 700 via ROM 702 and / or communication unit 709. When the computer program is loaded into RAM 703 and executed by the computing unit 701, one or more steps of the model rendering method described above may be performed. Alternatively, in other embodiments, the computing unit 701 may be configured to perform the model rendering method by any other suitable means (e.g., by means of firmware).

[0086] Various embodiments 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-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0087] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0088] In the context of this disclosure, a machine-readable medium can 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 can be, 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 include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0089] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0090] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0091] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, distributed system servers, or servers incorporating blockchain technology.

[0092] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution provided in this disclosure can be achieved, and this is not limited herein.

[0093] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A model rendering method, comprising: Learn the first-precision white model in the first engine; Learn the second-precision multi-angle texture data in the second engine, where the second precision is higher than the first precision; The multi-angle texture data is mapped onto the white model to obtain the rendered model; The second-precision multi-angle texture data learned in the second learning engine includes: The reference rendering model in the second engine is extracted from different angles to obtain reference multi-angle texture data; The reference multi-angle texture data and the white model are fused and trained to obtain the multi-angle texture data; The step of fusing and training the reference multi-angle texture data and the white model to obtain the multi-angle texture data includes: Determine at least one of the target lighting, target viewpoint, and target pose texture space of the white model; The reference multi-angle texture data is set to be the same as at least one of the target lighting, target viewpoint and target pose texture space to generate the multi-angle texture data; The step of mapping the multi-angle texture data onto the white model to obtain the rendered model includes: The white model and the multi-angle texture data are fitted together to generate rendering data in the first engine that couples the texture with the white model. The rendering data is input into the first engine to render the white model, thereby obtaining the rendered model.

2. The method according to claim 1, wherein, The step of inputting the rendering data into the first engine to render the white model and obtain the rendered model includes: The rendering data is input into a facial wireless mesh network, and facial expression driving, bone driving, and unfolding coloring are performed on the face of the white model in sequence to obtain a facial rendering model. The facial rendering model is input into the overall wireless mesh network, and facial expression driving, bone driving, and physically based shading and no-light shading are performed on the facial rendering model in sequence to obtain the rendering model.

3. A model rendering apparatus, comprising: The first learning module is configured to learn the first-precision white model in the first engine; The second learning module is configured to learn multi-angle texture data of a second precision from the second engine, wherein the second precision is higher than the first precision. The rendering module is configured to map the multi-angle texture data onto the white model to obtain the rendered model; The second learning module includes: The extraction submodule is configured to extract reference rendering models from the second engine from different angles to obtain reference multi-angle texture data; The fusion submodule is configured to fuse the reference multi-angle texture data and the white model for training to obtain the multi-angle texture data. The fusion submodule is further configured to: Determine at least one of the target lighting, target viewpoint, and target pose texture space of the white model; The reference multi-angle texture data is set to be the same as at least one of the target lighting, target viewpoint and target pose texture space to generate the multi-angle texture data; The rendering module includes: The fitting submodule is configured to fit the white model and the multi-angle texture data to generate rendering data in the first engine that couples the texture with the white model. The rendering submodule is configured to input the rendering data into the first engine to render the white model and obtain the rendered model.

4. The apparatus according to claim 3, wherein, The rendering submodule is further configured to: The rendering data is input into a facial wireless mesh network, and facial expression driving, bone driving, and unfolding coloring are performed on the face of the white model in sequence to obtain a facial rendering model. The facial rendering model is input into the overall wireless mesh network, and facial expression driving, bone driving, and physically based shading and no-light shading are performed on the facial rendering model in sequence to obtain the rendering model.

5. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-2.

6. A non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-2.

7. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-2.

Images