WO2023273113A1

WO2023273113A1 - Method and apparatus for generating expression model, and device, program and readable medium

Info

Publication number: WO2023273113A1
Application number: PCT/CN2021/132537
Authority: WO
Inventors: 刘森
Original assignee: 完美世界(北京)软件科技发展有限公司
Priority date: 2021-06-30
Filing date: 2021-11-23
Publication date: 2023-01-05
Also published as: CN113470149B; CN113470149A; CN114913278A

Abstract

Disclosed in the present invention are a method and apparatus for generating an expression model, and a device, a program and a readable medium. The method comprises: adding, to an initial character model, a first deformation target corresponding to an expression model to be transferred, so as to obtain a first expression model, wherein the initial character model is a model without an expression; acquiring first boundary data corresponding to a first joint area of the first expression model; if the first boundary data indicates that the first joint area includes at least two boundaries, setting the first joint area to be in a non-deformation state, and then reversing the first expression model to obtain an intermediate expression model; and adding, to the initial character model without an expression, a second deformation target corresponding to the intermediate expression model, so as to obtain a second expression model. By means of the present application, a redundant action at a joint between an expression part and a non-expression part can be removed, thereby facilitating an improvement in the realness of a character making an expression.

Description

Expression model generation method, device, equipment, program and readable medium

cross reference

This application claims the priority of the Chinese patent application submitted on June 30, 2021, with the application number "202110743217.1", and the title of the invention is "Method and device for generating expression model, storage medium, and computer equipment", the entire content of which is incorporated by reference incorporated in this application.

technical field

The present application relates to the technical field of animation production, in particular to a method, device, equipment, program and readable medium for generating an expression model.

Background technique

In the prior art, when transferring the deformation target of the template expression to the new character model, the expression part of the character is transferred to the new character model as a separate deformation target, and the non-expression parts are not transferred to simplify the expression binding. The purpose of saving resource consumption. However, the expression part and the non-expression part are connected, and only the expression part is transferred. For example, if the template expression corresponding to the head and neck is transferred to the new character model as the deformation target, the movement of the neck of the template expression will be substituted into the In the new character model, this may cause seams where the neck connects to the body when the new character makes expressions, reducing the authenticity of the new characters' expressions.

Contents of the invention

The present invention proposes the following technical solutions to overcome or at least partially solve or slow down the above-mentioned problems:

According to one aspect of the present invention, a method for generating an expression model is provided, including: adding a first deformation target corresponding to the expression model to be transferred to the initial character model to obtain a first expression model, and the initial character model is an expressionless model ; Obtain the first boundary data corresponding to the first connection area of the first expression model; if the first boundary data indicates that the first connection area contains at least two boundaries, then set the first connection area to not After the deformed state, reverse engineer the first expression model to obtain an intermediate expression model; and add a second deformation target corresponding to the intermediate expression model to the expressionless initial character model to obtain a second expression model.

According to another aspect of the present invention, an expression model generation device is provided, including: a first model generation module, configured to add a first deformation target corresponding to the expression model to be transferred to the initial character model to obtain the first expression model, The initial character model is an expressionless model; the boundary data acquisition module is used to obtain the first boundary data corresponding to the first connection area of the first expression model; the first animation export module is used to if the first boundary The data indicates that the first joint area contains at least two boundaries, then after setting the first joint area to an undeformed state, the first expression model is reversed to obtain an intermediate expression model; and the second model generation module uses Then add the second deformation target corresponding to the intermediate expression model to the initial character model without expression to obtain a second expression model.

According to still another aspect of the present invention, a kind of computer equipment is provided, comprising memory, processor and computer program/instruction stored on the memory, when said processor executes said computer program/instruction, realizes the expression model provided by the present invention The steps of the generation method.

According to another aspect of the present invention, a computer-readable medium is provided, on which computer programs/instructions are stored, and when the computer programs/instructions are executed by a processor, the steps of the expression model generation method provided by the present invention are implemented.

According to still another aspect of the present invention, a computer program product is provided, including computer programs/instructions, and when the computer program/instructions are executed by a processor, the steps of the method for generating an expression model provided by the present invention are implemented.

The beneficial effect of the present invention is: by means of the above-mentioned technical solution, the method, device, equipment, program and readable medium for generating an expression model provided by the application, when transferring expressions to the initial role model, first add the expression to the initial role model. The first deformation target corresponding to the expression model to be transferred obtains the first expression model, if the expression display part of the first expression model corresponds to the first connection area corresponding to the non-expression display part, there are multiple boundaries, then the first connection area is set to no After the deformation state, reverse the first expression model to obtain the intermediate expression model with the second deformation target, so as to remove the vertex displacement of the first connection area in the deformation target of the original first expression model, and finally use the intermediate expression model as the deformation The goal is to add a second expression model to the initial character model with no expression, so that there will be no seams in the joining area of the second expression model when making expressions. When transferring expressions to the new role model, this application can verify the connection area between the expression display part and the non-expression display part, and if there are multiple boundaries in the connection area, set the area to an undeformed state and Carry out the reverse operation of the model, so as to add the reverse result to the new character model without expression, so that the final model has a realistic and seamless effect when making expressions. expression transmission efficiency.

Description of drawings

These and various other advantages and benefits of the present invention will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. In the attached picture:

FIG. 1 shows a schematic flow diagram of a method for generating an expression model provided by an embodiment of the present application;

Fig. 2 shows a schematic diagram of a model connection area provided by an embodiment of the present application;

Fig. 3 shows a schematic diagram of boundary lines of a model joint area provided by an embodiment of the present application;

FIG. 4 shows a schematic structural diagram of a device for generating an expression model provided by an embodiment of the present application;

FIG. 5 shows a schematic structural diagram of a computer device provided by an embodiment of the present application;

FIG. 6 shows a schematic structural diagram of a computer program product provided by an embodiment of the present application.

detailed description

The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments. The following description is only to illustrate the basic principle of the present invention and not to limit it.

A method for generating an expression model is provided in this embodiment, as shown in Figure 1, the method includes:

Step 101, adding the first deformation target corresponding to the expression model to be transferred to the initial character model to obtain the first expression model, and the initial character model is an expressionless model;

In the embodiment of the present application, the initial role model is an expressionless model that needs to transmit expressions. The expressionless initial role model may include expression display parts and non-expression display parts, or may only include expression display parts. Among them, the expression display parts It can be the head of the model, or the head and neck of the model. The non-expression display part is the other part of the initial character model except the expression display part. The first deformation target is specifically the deformation target corresponding to the expression display part. That is, the deformation target of the expression model to be transferred. In a specific application scenario, the deformation target can be Blendshape in maya software (MAYA software is a famous 3D modeling and animation software owned by Autodesk), or 3dsmax (3D Studio Max, often referred to as 3d Max or 3ds MAX, It is a 3D animation rendering and production software based on PC system developed by Discreet (later merged by Autodesk). It is called Morph, and the deformation target can make the character model vertices occur according to the specified deformation target without binding the bones. displacement.

In the above-mentioned embodiment, taking the initial character model including expression display parts and non-expression display parts as an example, in order to realize expression transfer, the first deformation target of the expression model to be transferred is first added to the expression display parts of the initial character model without expression , the first expression model is obtained. The first expression model is the initial character model carrying the corresponding expression of the first deformation target. At this time, the expression display part of the first expression model has expression features, and the non-expression display part is still the same as the expressionless one. The initial character model is consistent.

Step 102, acquiring first boundary data corresponding to the first connection area of the first expression model;

Next, since the expression display part in the first expression model has the expression corresponding to the first deformation target, the non-expression display part is still consistent with the original character model without expression. In some cases, for example, the expression display part is the head and neck. When the expression corresponding to the first deformation target includes the neck deformation, if the model directly displays the expression, there may be seams at the connection between the neck and the body due to deformation. In order to determine whether the connection position between the expression display part and the non-expression display part is There will be gaps causing character models to display expressions that are not realistic. Or when the initial character model is a model of an expression display part, avoid gaps in the connection between the initial character model and other models when the initial character model is combined with other models that are not part of the expression display. In the embodiment of the present application, the first boundary data corresponding to the connection between the expression display part and the non-expression display part in the first expression model is obtained, that is, the first boundary data corresponding to the first connection area, and the expression display part is the head In the case of neck and neck, as shown in Figure 2, the first connection area can be a circle connected to the body and close to the lower neck, that is, the dark area in the figure, and the first boundary data is specifically the character model when making expressions For the boundary line generated under the action of deformation, if there are multiple boundary lines at the same position corresponding to the first joint area, it means that the deformation target of the deformation target has moved at the joint.

Optionally, step 102 may specifically include: performing reverse engineering on the expression display part of the first expression model to obtain the first boundary data, and displaying the first boundary data in the form of a wireframe view, so that the The model boundaries corresponding to the expression display parts of the first expression model are displayed in the form of lines.

In the above embodiment, the first boundary data can be obtained by performing a reverse operation on the first expression model, wherein the reverse operation specifically refers to copying each deformation target of the model as a separate model, and after obtaining the boundary data through the reverse operation, The boundary lines of the first connecting region can be displayed in the form of a wireframe view, as shown in FIG. 3 , and can be displayed in the form of a side view for convenience of observation. The reverse operation can be performed through the maya python script. First select the model to be reversed, and then run the script. The script function can be integrated in the expression workflow auxiliary toolkit. The script code can be as follows:

Step 103, if the first boundary data indicates that the first joint area contains at least two boundaries, then after setting the first joint area to a non-deformed state, reverse engineer the first expression model to obtain an intermediate expression model ;

In the embodiment of this application, if the same position corresponding to the first connecting area contains two or more boundary lines, it means that some facial expressions drive the movement of the model boundaries of the facial expression parts, which will cause differences with the non-expression display parts. Realistic seam, the same position can specifically refer to a model edge or a model surface corresponding to the expression display part in the model. In the actual application scenario, the tester can zoom in after the first boundary data is displayed in a wireframe view Observe to judge whether the human eyes can see the seam when the character model produces facial expressions, and also recognize the model boundary line through intelligent recognition, for example, the distance between the same position and the character model itself is greater than the preset Whether the number of morphing target boundary lines of the value (the morphing target boundary line refers to the boundary line corresponding to the first morphing target) is two or more. If it contains at least two boundaries, set the first connection area to the non-deformed state, that is, set the area as a vertex that will not be affected by the change of the morph target, and then reverse the first expression model to obtain the intermediate expression with the second morph target The model is to erase the displacement of the morphing target in the first connection area in the first morphing target to generate the second morphing target, and remove the morphing target that will cause an unreal seam between the expression display part and the non-expression display part.

Optionally, step 103 may specifically include: if the first boundary data indicates that the first converging region contains at least two boundaries, then setting the deformation weight corresponding to the first converging region to 0, so that the The first converging area is not affected by the deformation of the first morphing object; the intermediate expression model is obtained by reverse engineering the first expression model, wherein the second morphing object corresponding to the intermediate expression model does not include the first morphing object A morph target corresponding to an articulation area.

In the above embodiment, taking Maya as an example, you can specifically open the drawing blend shape weight tool, set the drawing operation of the drawing tool to replace, and change the value to 0, and then perform the reverse operation on the first expression model. After the reverse operation, it will be All the deformation targets of the model are obtained, that is, the intermediate expression model with the second deformation target is obtained.

Step 104, adding a second deformation target corresponding to the intermediate expression model to the initial character model without expression to obtain a second expression model.

Finally, reassign the second deformation target of the intermediate expression model to the initial role model without expression to obtain the second expression model, so that when the final obtained second expression model produces expression changes based on the second deformation target, the expression display parts and non-expression There will be no seams between the expression display parts, and the second expression model can display expressions more realistically. When the template expression is assigned to a new character model, not only can the template expression be basically retained, but also the expression display effect of the new character model can be guaranteed.

Optionally, step 104 may specifically include: adding a second deformation target corresponding to the intermediate expression model to the expressionless initial character model copy to obtain the intermediate expression model; or, deleting the first deformation target of the first expression model. After the morph object, a second morph object corresponding to the intermediate expression model is added to the initial character model without expression to obtain the second expression model.

In the above embodiment, the expressionless initial character model can be copied to obtain a copy of the initial character model, and the second deformation target can be added to the copy of the initial character model to obtain a second expression model, or the first expression model can be The historical operation record is deleted and restored to the character model without expression, and then the second deformation target is added to obtain the second expression model.

By applying the technical solution of this embodiment, when transferring expressions to the initial role model, first add the first deformation target corresponding to the expression model to be transferred to the initial role model to obtain the first expression model, if the expression display part of the first expression model is the same as The first cohesive area corresponding to the non-expression display part has multiple boundaries, then after setting the first cohesive area to the non-deformed state, reverse the first expression model to obtain the intermediate expression model with the second deformation target to remove the original There is the vertex displacement of the first connection area in the morph target of the first expression model, and finally the intermediate expression model is added as the morph target to the initial character model without expression to obtain the second expression model, so that the second expression model can make expressions There will be no seams in the joined area. In the embodiment of the present application, when the expression is transferred to the new character model, the connection area between the expression display part and the non-expression display part can be verified, and if there are multiple boundaries in the connection area, the area can be set as non-deformed State and perform model reverse operation, so as to add the reverse result to the new character model without expression, so that the final model has a realistic and seamless effect when making expressions, and at the same time, it only needs to pass the deformation target of the expression display part. , ensuring the efficiency of expression transmission.

In the embodiment of this application, optionally, the following steps may also be included:

Step 105, acquiring second boundary data corresponding to the second connection area of the second expression model;

Step 106, if the second boundary data indicates that the second connection area contains at least two boundaries, after setting the second connection area to a non-deformed state, based on the reverse result of the second expression model and the expressionless Updating the second facial expression model with the initial character model until the second connecting area corresponds to a boundary.

In the above-mentioned embodiment, after converting the first morph object into the second morph object and assigning the initial character model with no expression, in order to ensure the expression display effect, the second expression model can be verified, which can be compared with the first expression model. In a similar manner to the model, the second boundary data corresponding to the second connection area between the expression display part and the non-expression display part in the second expression model is obtained, so as to judge whether the second connection area contains multiple boundaries according to the second boundary data. Then set the second connection area to the non-deformed state, perform reverse operation on the second expression model and assign the reverse result to the initial character model without expression, and obtain a new second expression model until the same position of the second connection area contains only one up to the border.

Further, the embodiment of the present application may also include the following steps, which are performed before step 101 or separately, as follows:

Step 201, obtaining a third expression model and a fourth expression model, wherein the fourth expression model matches the model features of the initial character model, and the third expression model matches at least one model of the fourth expression model The initial states of the features are different, and the deformation target corresponding to the fourth expression model is obtained by modifying the vertex position corresponding to the at least one model feature in the third expression model;

Step 202, adding the fourth expression model as a deformation target to the third expression model to obtain a fifth expression model;

Step 203, after setting the at least one model feature of the fifth expression model to a deformed state, reverse the expression display part corresponding to the fifth expression model to obtain the expression model to be transferred, and the expression to be transferred The model corresponds to the first morph target.

In the above-mentioned embodiment, before the template expression is transferred, the template expression can be made first. In the embodiment of the present application, the third expression model and the fourth expression model are two models with different initial states of at least one model feature, for example The mouth characteristics of the third expression model and the fourth expression model are different, the third expression model is a model with an open mouth, and the fourth expression model is a model with a closed mouth. Optionally, step 201 may specifically include: obtaining a third expression model; Copying the third expression model, and adjusting the vertex position of the at least one model feature corresponding to the third expression model, to obtain the fourth expression model. Specifically, the third expression model can be copied to obtain the fourth expression model, and then the mouth can be closed without destroying the topological relationship. Specifically, the fourth expression model can be changed from open mouth to closed mouth by adjusting the position of the vertices of the model. Then, the fourth expression model is added to the third expression model as the deformation target, so that the closed mouth expression model becomes the deformation target and added to the mouth opening expression model, and the fifth expression with the initial state of opening the mouth and the deformation target of closing the mouth is obtained model, and finally set at least one of the above-mentioned model features to a deformed state, for example, set the mouth closure value of the fifth expression model to 1, reverse the fifth expression model, and obtain the first deformation target whose deformation target is closed mouth, Then add the first deformation target to the fourth expression model, without the need for the modeler to redraw the deformation target of the shut-up model, the animation of the mouth-opening model can be changed to the animation of the shut-up model, which is simple and convenient.

Further, as a specific implementation of the method in Figure 1, the embodiment of the present application provides an expression model generation device, as shown in Figure 4, the device includes:

The first model generation module is used to add the first deformation target corresponding to the expression model to be transferred to the initial role model to obtain the first expression model, and the initial role model is an expressionless model;

A boundary data acquisition module, configured to acquire first boundary data corresponding to the first connection area of the first expression model;

The first animation exporting module is configured to, if the first boundary data indicates that the first connection region contains at least two boundaries, set the first connection region to a non-deformed state, and then perform an operation on the first expression model. Get the intermediate expression model in reverse;

The second model generating module is configured to add a second deformation target corresponding to the intermediate expression model to the initial character model without expression, to obtain a second expression model.

Optionally, the boundary data acquisition module is specifically configured to: perform reverse engineering on the expression display part of the first expression model to obtain the first boundary data, and display the first boundary data in the form of a wireframe view , so that the model boundaries corresponding to the expression display parts of the first expression model are displayed in the form of lines.

Optionally, the first animation exporting module is specifically configured to: set the deformation weight corresponding to the first connection area to 0, so that the first connection area is not affected by the deformation of the first deformation target ; performing reverse engineering on the first expression model to obtain the intermediate expression model, wherein the second morph object corresponding to the intermediate expression model does not include the morph object corresponding to the first connecting region.

Optionally, the second model generation module is specifically configured to: add a second deformation target corresponding to the intermediate expression model to the expressionless initial character model copy to obtain the intermediate expression model; or, delete the first After the first morph object of an expression model, add a second morph object corresponding to the intermediate expression model to the initial character model without expression, to obtain the second expression model.

Optionally, the device also includes:

The verification module is used to add the second deformation target corresponding to the intermediate expression model to the initial role model without expression, and after obtaining the second expression model, obtain the corresponding second connection area of the second expression model the second boundary data; if the second boundary data indicates that the second connection area contains at least two boundaries, after setting the second connection area to a non-deformed state, based on the reverse result of the second expression model And the initial character model without expression updates the second expression model until the second connecting area corresponds to a boundary.

Optionally, the device also includes:

The second animation export module is used to add the first deformation target corresponding to the expression model to be transferred to the initial character model, and obtain the third expression model and the fourth expression model before obtaining the first expression model, wherein the fourth The expression model is matched with the model features of the initial character model, the initial state of at least one model feature of the third expression model is different from that of the fourth expression model, and the deformation target corresponding to the fourth expression model is passed through the The vertex position corresponding to the at least one model feature in the third expression model is modified and obtained; the fourth expression model is added to the third expression model as a deformation target to obtain a fifth expression model; the fifth expression is set. After the at least one model feature of the model is in a deformed state, reverse the expression display part corresponding to the fifth expression model to obtain the expression model to be transferred, and the expression model to be transferred corresponds to the first deformation target .

It should be noted that, for other corresponding descriptions of the functional units involved in an expression model generating apparatus provided in the embodiment of the present application, reference may be made to the corresponding descriptions in the method in FIG. 1 , and details are not repeated here.

The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all functions of some or all components in the expression model generating apparatus according to the embodiment of the present invention. The present invention can also be implemented as programs/instructions (eg, computer programs/instructions and computer program products) of devices or means for performing part or all of the methods described herein. Such programs/instructions for implementing the present invention may be stored on a computer-readable medium, or may exist in the form of one or more signals, such signals may be downloaded from an Internet website, or provided on a carrier signal, or in any form Available in other formats.

Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can be implemented by any method or technology for storage of information. Information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Flash memory or other memory technology, Compact Disc Read-Only Memory (CD-ROM), Digital Versatile Disc (DVD) or other optical storage, Magnetic cassettes, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by computing devices.

FIG. 5 schematically shows a computer device that can implement the method for generating an expression model according to the present invention, and the computer device includes a processor 410 and a computer-readable medium in the form of a memory 420 . Memory 420 is one example of a computer readable medium having storage space 430 for storing computer programs/instructions 431 . When the computer program/instruction 431 is executed by the processor 410, various steps in the method for generating an expression model described above can be realized.

Fig. 6 schematically shows a block diagram of a computer program product implementing the method according to the invention. Described computer program product comprises computer program/instruction 510, when described computer program/instruction 510 is executed by processors such as processor 410 shown in Figure 5, can realize the generation method of expression model described above each step in the .

The foregoing describes certain embodiments of the specification which, together with other embodiments, are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily follow the particular order shown, or sequential order, to achieve desirable results. Multitasking and parallel processing are also possible or advantageous in certain embodiments.

It should also be noted that the term "comprises", "comprises" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus comprising a set of elements includes not only those elements, but also includes Other elements not expressly listed, or elements inherent in the process, method, commodity, or apparatus are also included. Without further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising said element.

It should be understood that the above-mentioned embodiments are only for the purpose of illustrating the present invention rather than limiting the present invention. Without departing from the basic spirit and characteristics of the present invention, those skilled in the art can implement the present invention in other ways. The scope of the present invention should be determined by the appended claims, and any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of one or more embodiments of the present specification shall be covered therein.

Claims

A method for generating an expression model, comprising:

Adding the first deformation target corresponding to the expression model to be transferred to the initial role model to obtain the first expression model, the initial role model is an expressionless model;

Acquiring first boundary data corresponding to the first connection area of the first expression model;

If the first boundary data indicates that the first joint area contains at least two boundaries, after setting the first joint area to an undeformed state, reverse the first expression model to obtain an intermediate expression model; and

A second deformation target corresponding to the intermediate expression model is added to the initial character model without expression to obtain a second expression model.
The method according to claim 1, wherein the acquiring the first boundary data corresponding to the first connection area of the first expression model specifically comprises:

Reversing the expression display part of the first expression model to obtain the first boundary data, and displaying the first boundary data in the form of a wireframe view, so that the expression display part of the first expression model corresponds to Model boundaries are displayed as lines.
The method according to claim 1, characterized in that, after setting the first connecting region to an undeformed state, performing reverse engineering on the first expression model to obtain an intermediate expression model, specifically comprising:

Setting the deformation weight corresponding to the first converging area to 0, so that the first converging area is not affected by the deformation of the first deformation target;

Reversing the first expression model to obtain the intermediate expression model, wherein the second morphing object corresponding to the intermediate expression model does not include the morphing object corresponding to the first connecting region.
The method according to claim 1, wherein said adding a second deformation target corresponding to said intermediate expression model to said initial character model without expression, to obtain a second expression model, specifically comprising:

adding a second deformation target corresponding to the intermediate expression model to the expressionless initial character model copy to obtain the intermediate expression model; or,

After deleting the first morph object of the first expression model, adding a second morph object corresponding to the intermediate expression model to the initial character model without expression, to obtain the second expression model.
The method according to claim 1, characterized in that, adding the second deformation target corresponding to the intermediate expression model to the initial character model without expression, and after obtaining the second expression model, the method also includes:

Acquiring second boundary data corresponding to the second connection area of the second expression model;

If the second boundary data indicates that the second connection area contains at least two boundaries, after setting the second connection area to the non-deformed state, based on the reverse result of the second expression model and the expressionless initial character The model updates the second expression model until the second connecting area corresponds to a boundary.
The method according to any one of claims 1 to 5, wherein the first deformation target corresponding to the expression model to be transferred is added to the initial character model, and before the first expression model is obtained, the method also includes:

Acquiring a third expression model and a fourth expression model, wherein the fourth expression model matches the model features of the initial role model, and the third expression model matches the initial character of at least one model feature of the fourth expression model The states are different, and the deformation target corresponding to the fourth expression model is obtained by modifying the vertex position corresponding to the at least one model feature in the third expression model;

Adding the fourth expression model as a deformation target to the third expression model to obtain a fifth expression model;

After setting the at least one model feature of the fifth expression model to a deformed state, reverse the expression display part corresponding to the fifth expression model to obtain the expression model to be transferred, and the expression model to be transferred corresponds to The first morph target.
A device for generating an expression model, comprising:

The first model generation module is used to add the first deformation target corresponding to the expression model to be transferred to the initial role model to obtain the first expression model, and the initial role model is an expressionless model;

A boundary data acquisition module, configured to acquire first boundary data corresponding to the first connection area of the first expression model;

The first animation exporting module is configured to, if the first boundary data indicates that the first connection region contains at least two boundaries, set the first connection region to a non-deformed state, and then perform an operation on the first expression model. Reversely obtain the intermediate expression model; and

The second model generating module is configured to add a second deformation target corresponding to the intermediate expression model to the initial character model without expression, to obtain a second expression model.
The device according to claim 7, characterized in that,

The boundary data acquisition module is specifically configured to: obtain the first boundary data by reverse engineering the expression display part of the first expression model, and display the first boundary data in the form of a wireframe view, so that all The model boundaries corresponding to the expression display parts of the first expression model are displayed in the form of lines.
A computer device comprising a memory, a processor and computer programs/instructions stored on the memory, when the processor executes the computer program/instructions, the expression model according to any one of claims 1-6 is realized The steps to generate the method.
A computer-readable medium, on which computer programs/instructions are stored, and when the computer programs/instructions are executed by a processor, the steps of the method for generating an expression model according to any one of claims 1-6 are implemented.
A computer program product, including computer programs/instructions, when the computer programs/instructions are executed by a processor, the steps of the method for generating an expression model according to any one of claims 1-6 are realized.