CN115713582B

CN115713582B - Avatar generation method, device, electronic equipment and medium

Info

Publication number: CN115713582B
Application number: CN202211546700.1A
Authority: CN
Inventors: 彭昊天; 陈睿智; 赵晨
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2022-12-02
Filing date: 2022-12-02
Publication date: 2023-10-27
Anticipated expiration: 2042-12-02
Also published as: CN115713582A

Abstract

The present disclosure provides a method and an apparatus for generating an avatar, which relate to the field of artificial intelligence, in particular to the fields of augmented reality, virtual reality, computer vision, deep learning, etc., and can be applied to scenes such as meta universe, virtual digital people, etc. The specific implementation scheme is as follows: acquiring a reference initial posture parameter for a reference avatar and a target initial posture parameter for a target avatar; adjusting the initial pose parameter of the target based on the initial pose parameter of the reference so that the initial pose of the adjusted target avatar matches the initial pose of the reference avatar; acquiring a target attitude parameter associated with a target attitude of a reference virtual image, and acquiring an attitude adjustment parameter according to the target attitude parameter and a reference initial attitude parameter; and generating the target virtual image based on the target posture parameter according to the posture adjustment parameter and the adjusted target initial posture parameter of the target virtual image.

Description

Avatar generation method, device, electronic equipment and medium

Technical Field

The disclosure relates to the technical field of artificial intelligence, in particular to the fields of augmented reality, virtual reality, computer vision, deep learning and the like, and can be applied to scenes such as metauniverse, virtual digital people and the like. The present disclosure relates in particular to an avatar generation method, apparatus, electronic device, storage medium and computer program product.

Background

Action redirection refers to the action of a given avatar, migrating that action to another avatar, keeping its actions consistent. Action redirection has wide application in social, live, gaming, etc. scenarios. In the related art, the action redirection may be accomplished in an offline manner. However, in practice the acquisition of an avatar as a source of motion is usually real-time and the way of off-line redirection is not adapted. Moreover, due to the multiple differences of bones, skins, patches, etc. between different avatars, adopting an off-line redirection approach typically requires a lot of manual debugging to iterate the configuration file, which can be time consuming.

Disclosure of Invention

The present disclosure provides an avatar generation method, apparatus, electronic device, storage medium, and computer program product.

According to an aspect of the present disclosure, there is provided an avatar generation method including: acquiring a reference initial posture parameter for a reference avatar and a target initial posture parameter for a target avatar; the initial posture parameter of the reference is used for representing the initial posture of the reference virtual image, and the initial posture parameter of the target is used for representing the initial posture of the target virtual image; adjusting the initial pose parameter of the target based on the initial pose parameter of the reference so that the initial pose of the adjusted target avatar matches the initial pose of the reference avatar; acquiring a target attitude parameter associated with a target attitude of a reference virtual image, and acquiring an attitude adjustment parameter according to the target attitude parameter and a reference initial attitude parameter; and generating the target virtual image based on the target posture parameter according to the posture adjustment parameter and the adjusted target initial posture parameter of the target virtual image.

According to another aspect of the present disclosure, there is provided an avatar generating apparatus including: a first acquisition module for acquiring a reference initial pose parameter for a reference avatar and a target initial pose parameter for a target avatar; the initial posture parameter of the reference is used for representing the initial posture of the reference virtual image, and the initial posture parameter of the target is used for representing the initial posture of the target virtual image; the adjusting module is used for adjusting the initial posture parameters of the target based on the initial posture parameters of the reference, so that the initial posture of the adjusted target virtual image is matched with the initial posture of the reference virtual image; the second acquisition module is used for acquiring target attitude parameters related to the target attitude of the reference virtual image and acquiring attitude adjustment parameters according to the target attitude parameters and the reference initial attitude parameters; and the generating module is used for generating the target virtual image based on the target posture parameter according to the posture adjustment parameter and the target initial posture parameter of the adjusted target virtual image.

According to another aspect of the present disclosure, there is provided an electronic device including: 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 to enable the at least one processor to perform a method provided in accordance with 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 perform a 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 a method provided according to the present disclosure.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the disclosure, nor is it intended to be used to limit the scope of the disclosure. Other features of the present disclosure will become apparent from the following specification.

Drawings

The drawings are for a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:

FIG. 1 is a schematic diagram of an exemplary system architecture to which avatar generation methods and apparatuses may be applied, according to embodiments of the present disclosure;

fig. 2 is a flowchart of an avatar generation method according to an embodiment of the present disclosure;

FIG. 3 is a schematic diagram of a tree hierarchy of skeletal nodes, according to an embodiment of the present disclosure;

fig. 4A and 4B are effect diagrams of a bone structure corresponding to an initial pose of a reference avatar and a bone structure corresponding to an initial pose of an adjusted target avatar, respectively;

Fig. 5A and 5B are effect graphs of a target pose of a reference avatar and a target pose of a target avatar based on a target pose parameter, respectively;

fig. 6 is a flowchart of an avatar generation method according to another embodiment of the present disclosure;

fig. 7 is a block diagram of an avatar generating apparatus according to an embodiment of the present disclosure; and

fig. 8 is a block diagram of an electronic device for implementing an avatar generation method of an embodiment of the present disclosure.

Detailed Description

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 should be considered as merely exemplary. Accordingly, one 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 present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and/or the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It should be noted that the terms used herein should be construed to have meanings consistent with the context of the present specification and should not be construed in an idealized or overly formal manner.

Where expressions like at least one of "A, B and C, etc. are used, the expressions should generally be interpreted in accordance with the meaning as commonly understood by those skilled in the art (e.g.," a system having at least one of A, B and C "shall include, but not be limited to, a system having a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).

The embodiment of the disclosure provides an avatar generation method, comprising the following steps: acquiring a reference initial posture parameter for a reference avatar and a target initial posture parameter for a target avatar; adjusting the initial pose parameter of the target based on the initial pose parameter of the reference so that the initial pose of the adjusted target avatar matches the initial pose of the reference avatar; acquiring a target attitude parameter associated with a target attitude of a reference virtual image, and acquiring an attitude adjustment parameter according to the target attitude parameter and a reference initial attitude parameter; and generating the target virtual image based on the target posture parameter according to the posture adjustment parameter and the adjusted target initial posture parameter of the target virtual image.

Fig. 1 is a schematic diagram of an exemplary system architecture to which avatar generation and apparatuses may be applied according to an embodiment of the present disclosure. It should be noted that fig. 1 is only an example of a system architecture to which embodiments of the present disclosure may be applied to assist those skilled in the art in understanding the technical content of the present disclosure, but does not mean that embodiments of the present disclosure may not be used in other devices, systems, environments, or scenarios.

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 to provide communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired and/or wireless communication links, and the like.

The user may interact with the server 105 via the network 104 using the terminal devices 101, 102, 103 to receive or send messages or the like. Various client applications can be installed on the terminal devices 101, 102, 103. For example, an animation class application, a live class application, a game class application, a web browser application, a search class application, an instant messaging tool, a mailbox client or social platform software, and the like (just examples).

The terminal devices 101, 102, 103 may be a variety of electronic devices having a display screen and supporting web browsing, including but not limited to smartphones, tablets, laptop and desktop computers, and the like.

The server 105 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as cloud computing, network service, and middleware service.

The server 105 may be a server providing various services, such as a background management server (by way of example only) providing support for websites browsed by users using the terminal devices 101, 102, 103. The background management server may analyze and process the received data such as the user request, and feed back the processing result (e.g., the web page, information, or data obtained or generated according to the user request) to the terminal device.

For example, the server 105 may acquire a reference initial pose parameter for the reference avatar and a target initial pose parameter for the target avatar from the terminal devices 101, 102, 103 through the network 104, and adjust the target initial pose parameter based on the reference initial pose parameter such that the initial pose of the adjusted target avatar matches the initial pose of the reference avatar. The method comprises the steps of obtaining target posture parameters related to target postures of a reference virtual image, obtaining posture adjustment parameters according to the target posture parameters and the reference initial posture parameters, and generating a target virtual image based on the target posture parameters according to the posture adjustment parameters and the adjusted target initial posture parameters of the target virtual image. The server 105 may also transmit the target avatar based on the target pose parameter to the terminal devices 101, 102, 103.

It should be noted that the avatar generation method provided by the embodiments of the present disclosure may be generally performed by the server 105. Accordingly, the avatar generating apparatus provided by the embodiments of the present disclosure may be generally provided in the server 105. The avatar generation method provided by the embodiments of the present disclosure may also be performed by a server or a server cluster that is different from the server 105 and is capable of communicating with the terminal devices 101, 102, 103 and/or the server 105. Accordingly, the avatar generating apparatus provided by the embodiments of the present disclosure may also be provided in a server or server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103 and/or the server 105.

It should be understood that the number of terminal devices, networks and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

It should be noted that the sequence numbers of the respective operations in the following methods are merely representative of the operations for the purpose of description, and should not be construed as representing the order of execution of the respective operations. The method need not be performed in the exact order shown unless explicitly stated.

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

As shown in fig. 2, the avatar generation method 200 may include operations S210 to S240, for example.

In operation S210, a reference initial pose parameter for a reference avatar and a target initial pose parameter for a target avatar are acquired.

In operation S220, the target initial pose parameter is adjusted based on the reference initial pose parameter such that the initial pose of the adjusted target avatar matches the initial pose of the reference avatar.

In operation S230, a target pose parameter associated with a target pose of the reference avatar is acquired, and a pose adjustment parameter is obtained according to the target pose parameter and the reference initial pose parameter.

In operation S240, a target avatar based on the target pose parameter is generated according to the pose adjustment parameter and the adjusted target initial pose parameter of the target avatar.

According to an embodiment of the present disclosure, the target virtual image is a virtual object to be redirected. The target avatar may be subjected to an action redirection process using the reference avatar as an action source so as to migrate the posture of the reference avatar onto the target avatar such that the posture of the target avatar coincides with the posture of the reference avatar. It will be appreciated that the pose of the avatar is determined by the skeletal architecture of the avatar, and thus, the reference avatar is used as a motion source to perform a motion redirection process on the target avatar, essentially migrating the skeletal architecture of the reference avatar onto the target avatar.

The avatar may be a character in a video such as a game or an animation, or may be another suitable character, and may specifically be selected according to an actual application scenario. The avatar may be in the form of a character, a cartoon, or other forms, which is not limited by the present disclosure.

According to the embodiment of the present disclosure, the reference avatar and the target avatar having initial poses may be respectively input into the deep learning model, resulting in reference initial pose parameters for the reference avatar and target initial pose parameters for the target avatar. It will be appreciated that the deep learning model may be any deep learning model for human body posture estimation, for example, but not limited to, a SMPL (skyned Multi-Person Linear Model) series model, and may be specifically selected according to the actual application scenario.

In the disclosed embodiments, a reference initial pose parameter for a reference avatar may be used to characterize an initial pose of the reference avatar, and a target initial pose parameter for a target avatar may be used to characterize an initial pose of the target avatar.

It can be appreciated that since there are various differences in bones, skins, facings, etc. between different avatars, there may be differences in the initial pose of the reference avatar from the initial pose of the target avatar. In addition, for the same avatar in different poses, for example, reference avatars having different poses, their corresponding bone architectures are also different. Moreover, since the acquisition of the target pose of the reference avatar is real-time in practice, if the target pose of the reference avatar is directly utilized to redirect the target avatar, the pose parameters of the avatar need to be manually and continuously debugged, which not only generates great time consumption, but also affects the accuracy of the avatar redirecting result.

Accordingly, the initial pose of the target avatar may be first adjusted to coincide with the initial pose of the reference avatar. In this way, the redirection of the target avatar may be accomplished subsequently by migrating a difference between the target pose of the reference avatar and the initial pose of the reference avatar into the target avatar. By adopting the mode, the artificial auxiliary workload in the action redirection process of the virtual image is reduced, and the efficiency and the accuracy of the action redirection process can be improved.

For example, the target initial pose parameter may be adjusted based on the reference initial pose parameter such that the adjusted initial pose of the target avatar matches the initial pose of the reference avatar.

Next, a target pose parameter associated with a target pose of the reference avatar may be acquired, and a pose adjustment parameter may be obtained based on the target pose parameter and the reference initial pose parameter. And then, generating the target virtual image based on the target posture parameters according to the posture adjustment parameters and the adjusted target initial posture parameters of the target virtual image.

It should be noted that, the manner of acquiring the target posture parameter for the reference avatar is similar to the manner of acquiring the target initial posture parameter for the target avatar. For example, a reference avatar having a target pose may be input into the deep learning model described above, resulting in target pose parameters for the reference avatar.

In addition, the target pose parameters for the reference avatar may include target pose parameters of the reference avatar at a plurality of target poses. Wherein the reference avatar in each target pose corresponds to a target pose parameter for the reference avatar. The above operations S2 to S1 0 to S240 may be employed for the target pose parameters of the reference avatar at each target pose to generate a target avatar based on the target pose parameters. Thus, the reference avatar in the plurality of target poses can be redirected onto the target avatar, resulting in the target avatar in the plurality of target poses. Based on the mode, the requirement of real-time redirection can be met, and the efficiency of action redirection processing is improved.

By the embodiments of the present disclosure, the target initial pose parameter of the target avatar is adjusted based on the reference initial pose parameter for the reference avatar such that the adjusted initial pose of the target avatar matches the initial pose of the reference avatar. And then, processing the target initial posture parameters of the adjusted target virtual image based on the posture difference between the target posture and the initial posture of the reference virtual image to obtain a target virtual image based on the target posture parameters, so as to realize the migration of the target posture of the reference virtual image to the target virtual image, namely, the action redirection processing of the target virtual image is completed. The scheme disclosed by the invention is beneficial to reducing the manual auxiliary workload in the action redirection process of the virtual image, and can improve the efficiency of the action redirection processing and the accuracy of the action redirection processing.

For convenience of the following description, the contents related to the skeletal nodes of the avatar will be first described as an example.

According to an embodiment of the present disclosure, the skeleton architecture of the avatar includes a plurality of skeleton nodes, the skeleton nodes in the avatar have a topological association relationship therebetween, and a topological hierarchy formed by the plurality of skeleton nodes may be a tree hierarchy, for example.

For example, in a tree hierarchy of skeletal nodes, a previous skeletal node associated with a current skeletal node may be considered a parent node of the current skeletal node, and a next skeletal node associated with the current skeletal node may be considered a child node of the current skeletal node. The topology structure hierarchy corresponding to the father node of the current skeleton node is higher than the topology structure hierarchy corresponding to the current skeleton node, and the topology structure hierarchy corresponding to the current skeleton node is higher than the topology structure hierarchy corresponding to the child node of the current skeleton node.

The tree hierarchy of skeletal nodes of the present embodiment is illustrated below with reference to fig. 3.

Fig. 3 is a schematic diagram of a tree hierarchy of skeletal nodes in accordance with an embodiment of the present disclosure.

As shown in fig. 3, the skeleton node nodemot is a root node, the root node nodemot is a parent node of the skeleton node nodeA, and the skeleton node nodeA is a parent node of the skeleton node nodeB.

For example, a global pose matrix for a skeletal node may be determined using equation (1).

corrdinate _global ＝corrdinate _{global_parent} *corrdinate _local (1)

Corridinate in formula (1) _global Global pose matrix representing skeletal nodes _local Local pose matrix representing skeletal nodes, corridinate _{global_parent} A global pose matrix representing a parent node of a skeletal node.

The local gesture matrix of the skeleton node can indicate the relative position relation between the skeleton node and the father node, and the local gesture matrix of the father node can be transferred layer by layer from the node behind the root node to obtain the global gesture matrix of each skeleton node, namely the absolute position information of each skeleton node is obtained.

For example, the global pose matrices of skeletal nodes nodroot, skeletal nodes nodeA, and skeletal nodes nodeB may be determined using equation (2), equation (3), and equation (4), respectively.

globalRoot＝localRoot (2)

globalA＝globalRoot*localA (3)

globalB＝globalA*localB (4)

In equations (2) - (4), globalRoot, globalA, globalB represents the global pose matrix of bone node nodeRoot, nodeA, nodeB, respectively, and localRoot, localA, localB represents the local pose matrix of bone node nodeRoot, nodeA, nodeB, respectively.

According to an embodiment of the present disclosure, the reference initial pose parameter may include a first initial local pose matrix of a plurality of reference skeletal nodes of the reference avatar, and the target initial pose parameter may include a second initial local pose matrix of a plurality of target skeletal nodes of the target avatar.

In operation S220 described above, adjusting the target initial pose parameter based on the reference initial pose parameter may include the following operations.

For each of the second initial local pose matrices of the plurality of target bone nodes, adjusting the second initial local pose matrix based on the first initial local pose matrix corresponding to the second initial local pose matrix to obtain a third initial local pose matrix of the target bone node; and obtaining the target initial posture parameters of the adjusted target virtual image according to the plurality of third initial local posture matrixes.

For example, a first global pose matrix of the reference skeletal node may be determined from the first initial local pose matrix. And then, adjusting the second initial local gesture matrix according to the first global gesture matrix of the reference skeleton node, so as to obtain a third initial local gesture matrix of the target skeleton node.

The process of acquiring the third initial local pose matrix will be described below with reference to examples.

For example, for a root node rRoot in the reference avatar, a first global pose matrix for the root node rRoot may be determined from a first initial local pose matrix corresponding to the root node rRoot. And then, taking the first global posture matrix of the root node TRoot as a global posture matrix corresponding to the root node TRoot in the target virtual image, and determining a third initial local posture matrix corresponding to the root node TRoot according to the global posture matrix corresponding to the root node TRoot. As can be seen from formulas (1) to (4), the third initial local pose matrix of the root node TRoot is the same as the first initial local pose matrix corresponding to the root node rRoot.

Next, for the 2 nd reference skeleton node in the reference avatar, a first global pose matrix of the 2 nd reference skeleton node may be obtained according to the first initial local pose matrix of the 2 nd reference skeleton node and the first global pose matrix of the parent node of the 2 nd reference skeleton node (i.e., the first global pose matrix of the root node rRoot). And then, taking the first global gesture matrix of the 2 nd reference skeleton node as the global gesture matrix corresponding to the 2 nd target skeleton node in the target avatar. And then, according to the global gesture matrix corresponding to the 2 nd target skeleton node and the global gesture matrix corresponding to the root node TRoot, obtaining a third initial local gesture matrix corresponding to the 2 nd target skeleton node. It will be appreciated that since the first global pose matrix of the root node rRoot is the same as the global pose matrix corresponding to the root node TRoot, the third initial local pose matrix corresponding to the 2 nd target bone node is the same as the first initial local pose matrix of the 2 nd reference bone node.

Next, for an ith (i is an integer greater than 1) reference skeleton node in the reference avatar, a first global pose matrix of the ith reference skeleton node may be obtained from the first initial local pose matrix of the ith reference skeleton node and the first global pose matrix of the parent node of the ith reference skeleton node. Then, the first global posture matrix of the ith reference skeleton node is used as the global posture matrix of the ith target skeleton node. And then, according to the global gesture matrix of the ith target skeleton node and the global gesture matrix corresponding to the father node of the ith target skeleton node, obtaining a third initial local gesture matrix corresponding to the ith target skeleton node. Similarly, as can be seen from the above formulas (1) to (4), the third initial local pose matrix corresponding to the i-th target bone node is identical to the first initial local pose matrix of the i-th reference bone node.

As can be seen from the above description, the process of adjusting the initial pose parameters of the target based on the initial pose parameters of the reference is equivalent to replacing the second initial local pose matrix corresponding to each target bone node with the first initial local pose matrix corresponding to the corresponding reference bone node.

In some examples, adjusting the target initial pose parameter based on the reference initial pose parameter may also be accomplished by: for example, a preset gesture parameter corresponding to the preset gesture may be acquired, and then, the reference initial gesture parameter and the target initial gesture parameter of the target avatar may be adjusted based on the preset gesture parameter, respectively, such that the adjusted initial gesture of the reference avatar and the adjusted initial gesture of the target avatar are matched with the preset gesture, respectively.

According to embodiments of the present disclosure, preset gesture parameters are used to characterize a preset gesture (e.g., without limitation, a T-pos gesture) of any one avatar. The preset pose parameters include a preset local pose matrix of a plurality of skeletal nodes of the avatar. Based on the preset local pose matrix, the reference initial pose parameter and the target initial pose parameter of the target avatar may be adjusted. The process of adjusting the reference initial pose parameter and the target initial pose parameter of the target avatar based on the preset pose parameter is similar to the process of adjusting the target initial pose parameter based on the reference initial pose parameter, and is not repeated herein.

Through the above-described embodiments, the initial pose of the reference avatar and the initial pose (e.g., T-pos pose) of the target avatar may be adjusted in a preset pose of any one avatar. Thus, the initial posture of the adjusted target avatar can be made to coincide with the initial posture of the adjusted reference avatar. It should be noted that, in the embodiment of the present disclosure, the preset gesture may be set as required, which is not limited herein.

Fig. 4A and 4B are effect diagrams of a bone structure corresponding to an initial pose of a reference avatar and a bone structure corresponding to an initial pose of an adjusted target avatar, respectively. The effect of adjusting the initial pose of the target avatar based on the initial pose of the reference avatar is exemplarily described below with reference to fig. 4A and 4B.

In the embodiment of the present disclosure, according to the reference initial pose parameter of the reference avatar, a bone architecture corresponding to the initial pose of the reference avatar as shown in fig. 4A may be obtained. In addition, according to the above-described manner, the initial pose of the target avatar is adjusted based on the initial pose of the reference avatar, and a skeletal structure corresponding to the adjusted initial pose of the target avatar as shown in fig. 4B may be obtained.

It will be appreciated that the pose of the avatar is determined by the skeletal architecture of the avatar. Accordingly, the initial pose of the reference avatar and the initial pose of the adjusted target avatar can be determined according to the skeletal architecture of the avatars shown in fig. 4A and 4B, respectively.

As shown in fig. 4A, the initial pose of the reference avatar is, for example, a T-pos pose. As shown in fig. 4B, the initial pose of the adjusted target avatar is also a T-pos pose. As can be seen by comparing fig. 4A and 4B, the initial pose of the target avatar is adjusted based on the reference initial pose parameter of the reference avatar such that the adjusted initial pose of the target avatar is matched with the initial pose of the reference avatar.

In some embodiments, if the initial pose of the reference avatar and the initial pose of the target avatar are not T-pos poses and the initial poses of the two are not matched, the T-pos poses may be further used as preset poses, and the reference initial pose parameter and the target initial pose parameter of the target avatar are adjusted based on preset pose parameters corresponding to the preset poses, so that the adjusted initial pose of the target avatar is matched with the adjusted initial pose of the reference avatar.

According to an embodiment of the present disclosure, the target pose parameters include a first target local pose matrix of a plurality of reference skeletal nodes of the reference avatar, and deriving the pose adjustment parameters from the target pose parameters and the reference initial pose parameters may include the following operations. It should be noted that the following operations are performed for the first target local pose matrix of each reference skeletal node.

For example, for a first target local pose matrix for each reference skeletal node, a first pose adjustment coefficient for the reference skeletal node may be determined from the first target local pose matrix and the first initial local pose matrix. And then, according to the first initial local gesture matrix, obtaining a first global gesture matrix of the reference skeleton node, and according to the first gesture adjustment coefficient and the first global gesture matrix, determining gesture adjustment parameters.

For example, a first pose adjustment coefficient for a reference skeletal node may be determined according to equation (5).

Acorrdinate _{local_new} ＝Acorrdinate _{local_o1d} *ArotDiff _local (5)

In equation (5), acordinate _{local_new} First target local pose matrix representing reference skeletal node, acordnate _{local_old} A first initial local pose matrix representing a reference skeletal node, arotDiff _local Representing the first attitude adjustment factor.

In embodiments of the present disclosure, equation (6) may be employed to determine a first global pose matrix of reference skeletal nodes.

Acorrdinate _{global_o1d} ＝Acorrdinate _{global_parent_o1} d*Acorrdinate _{local_old} (6)

In equation (6), acordenate _{global_old} A first global pose matrix representing reference skeletal nodes, acordnate _{local_old} First initial local pose matrix representing reference skeletal nodes, acordnate _{global_parent_old} A first global pose matrix representing a parent node of a reference skeletal node.

For example, the attitude adjustment parameters may be determined according to formula (7).

ArotDiff _global ＝Acorrdinate _{global_old} *ArotDiff _local *Acorrdinate _{global_old} ^-1 (7)

In equation (7), arotDiff _global Representing attitude adjustment parameters, acordnate _{global_old} ^-1 An inverse of the first global pose matrix representing the reference skeletal node.

It should be noted that if the initial pose parameters of the reference avatar and the target initial pose parameters of the target avatar are adjusted based on the preset pose parameters, in determining the pose adjustment parameters, the first initial local pose matrix and the first global pose matrix corresponding to the reference skeleton node, which are referred to in the above formulas (5) to (7), represent the initial local pose matrix and the global pose matrix of the adjusted reference skeleton node of the reference avatar, respectively. The manner of determining the posture adjustment parameters based on the initial local posture matrix and the global posture matrix of the reference skeleton node of the adjusted reference avatar is similar to the above description, and will not be repeated here.

According to the embodiments of the present disclosure, after the posture adjustment parameters are obtained, the target avatar based on the target posture parameters may be generated according to the posture adjustment parameters and the target initial posture parameters of the adjusted target avatar. It should be noted that the following operations are performed for each third initial local pose matrix.

For example, for each third initial local pose matrix, a second global pose matrix for the target bone node may be derived from the third initial local pose matrix for the target bone node.

For example, a second global pose matrix of the target skeletal node may be determined according to equation (8).

Bcorrdinate _{global_o1d} ＝Bcorrdinate _{global_parent_old} *Bcorrdinate _{local_old} (8)

In equation (8), bcorrdinate _{global_old} A second global pose matrix representing the target skeletal node, bcrdininate _{local_old} A third initial local pose matrix representing the target bone node, bcrordinate _{global_parent_old} A second global pose matrix representing a parent node of the target skeletal node.

Next, a second pose adjustment coefficient may be determined based on the pose adjustment parameter and a second global pose matrix of the target skeletal node.

For example, the second attitude adjustment coefficient may be determined according to formula (9).

BrotDiff _local ＝Bcorrdinate _{global_o1d} *ArotDiff _global *Bcorrdinate _{global_o1d} ^-1 (9)

In equation (9), bcordininate _{global_o1d} A second global pose matrix representing the target skeletal node, bcrdininate _{global_o1d} ^-1 An inverse matrix, arotDiff, of a second global pose matrix representing the target skeletal node _global Representing attitude adjustment parameters, brotDiff _local Representing a second attitude adjustment factor.

Then, a second target local pose matrix of the target bone node may be obtained based on the second pose adjustment coefficient and the third initial local pose matrix of the target bone node.

For example, a second target local pose matrix for the target skeletal node may be determined according to equation (10).

Bcorrdinate _{local_new} ＝Bcorrdinate _{local_old} *BrotDiff _local (10)

In equation (10), bcorrdinate _{local_new} A second target local pose matrix representing target skeletal nodes, bcrdininate _{local_old} Representing a target skeletal nodeA third initial local pose matrix, brotDiff _local Representing a second attitude adjustment factor.

Next, a target avatar based on the target pose parameters may be obtained from the second target local pose matrix.

For example, the target pose parameters of the reference avatar may be converted into the target avatar by updating the third initial local pose of the target avatar using the second target local pose matrix of the target skeletal node, thereby generating the target avatar based on the target pose parameters.

In some embodiments, the target avatar based on the target pose parameters may also be generated in the following manner.

For example, if it is determined that the initial pose of the reference avatar matches the initial pose of the target avatar, that is, it may not be necessary to adjust the target initial pose parameter based on the reference initial pose parameter so that the adjusted initial pose of the target avatar matches the initial pose of the reference avatar. At this time, the target posture parameter associated with the target posture of the reference avatar may be directly acquired, and the posture adjustment parameter may be obtained according to the target posture parameter and the reference initial posture parameter. And then, generating the target virtual image based on the target posture parameters according to the posture adjustment parameters and the target initial posture parameters aiming at the target virtual image.

In the embodiment of the present disclosure, the posture adjustment parameter may be determined according to the target posture parameter of the reference avatar and the reference initial posture parameter of the reference avatar. The manner of determining the posture adjustment parameter is the same as that described above, and will not be described here again.

After determining the gesture adjustment parameters, a target avatar based on the target gesture parameters may be generated according to the gesture adjustment parameters and the target initial gesture parameters for the target avatar. It should be noted that the following operations are performed for the second initial local pose matrix of each target skeletal node.

For example, for each of the second initial local pose matrices of the plurality of target bone nodes, a third global pose matrix of the target bone node is derived from the second initial local pose matrices of the target bone nodes.

For example, a third global pose matrix of the target skeletal node may be determined according to equation (11).

Bcorrdinate _{global_old*} ＝Bcorrdinate _{global_parent_old*} *Bcorrdinate _{local_old*} (11)

In equation (11), bcordininate _{global_old*} A third global pose matrix, bcrdininate, representing a target skeletal node _{local_old*} A second initial local pose matrix representing the target bone node, bcrordinate _{global_parent_old*} A third global pose matrix representing a parent node of the target skeletal node.

Next, a third pose adjustment coefficient may be determined based on the pose adjustment parameter and a third global pose matrix of the target skeletal node.

For example, a third attitude adjustment factor may be determined according to equation (12).

BrotDiff _local* ＝Bcorrdinate _{global_old*} *ArotDiff _global *Bcorrdinate _{global_o1d*} ^-1 (12)

In equation (12), bcorrdinate _{global_old*} A third global pose matrix, bcrdininate, representing a target skeletal node _{global_old*} ^-1 An inverse matrix, arotDiff, of a third global pose matrix representing the target skeletal node _global Representing attitude adjustment parameters, brotDiff _local* Representing a third attitude adjustment factor.

And then, obtaining a third target local posture matrix of the target skeleton node according to the third posture adjustment coefficient and the second initial local posture matrix of the target skeleton node.

For example, a third target local pose matrix for the target skeletal node may be determined according to equation (13).

Bcorrdinate _{local_new*} ＝Bcorrdinate _{local_old*} *BrotDiff _local* (13)

In equation (13), bcordininate _{local_new*} A third target local pose matrix representing target skeletal nodes, bcrdininate _{local_o1d*} A second initial local pose matrix representing the target bone node, brotDiff _local* Representing a third attitude adjustment factor.

Next, a target avatar based on the target pose parameters is obtained according to the third target local pose matrix.

For example, the target pose parameters of the reference avatar may be converted into the target avatar by updating the second initial local pose of the target avatar with the third target local pose matrix of the target skeletal node, thereby generating the target avatar based on the target pose parameters.

Fig. 5A and 5B are effect diagrams of a target pose of a reference avatar and a target pose of a target avatar based on a target pose parameter, respectively. The effect of redirecting the target avatar based on the target pose of the reference avatar is illustrated below with reference to fig. 5A and 5B.

Fig. 5A illustrates an effect of a target pose of the reference avatar, and fig. 5B illustrates an effect of a target pose of the target avatar based on target pose parameters. As can be seen by comparing fig. 5A and 5B, the target pose of the reference avatar can be transferred to the target avatar by performing the redirection process based on the target pose of the reference avatar such that the pose of the target avatar is identical to the pose of the reference avatar.

Fig. 6 is a flowchart of an avatar generation method according to another embodiment of the present disclosure.

As shown in fig. 6, the avatar generation method 600 includes operations S601 to S608.

In operation S601, a reference initial pose parameter for a reference avatar is acquired.

In operation S602, a target initial pose parameter for a target avatar is acquired.

According to embodiments of the present disclosure, a reference initial pose parameter for a reference avatar may be used to characterize an initial pose of the reference avatar, and a target initial pose parameter for a target avatar may be used to characterize an initial pose of the target avatar.

The reference avatar and the target avatar having initial poses may be input into the deep learning model, respectively, to obtain reference initial pose parameters for the reference avatar and target initial pose parameters for the target avatar. The deep learning model is the same as the definition described above, and will not be described here again.

In operation S603, it is determined whether the initial pose of the reference avatar is matched with the initial pose of the target avatar, and if so, operation S605 is performed, otherwise, operation S604 is performed.

In operation S604, the target initial pose parameter is adjusted based on the reference initial pose parameter such that the initial pose of the adjusted target avatar matches the initial pose of the reference avatar.

Since there are various differences in bones, skins, facings, etc. between different avatars, there may be differences in the initial pose of the reference avatar and the initial pose of the target avatar. In addition, for the same avatar in different poses, for example, reference avatars having different poses, their corresponding bone architectures are also different. Moreover, since the acquisition of the target pose of the reference avatar is real-time in practice, if the target pose of the reference avatar is directly utilized to redirect the target avatar, the pose parameters of the avatar need to be manually and continuously debugged, which not only generates great time consumption, but also affects the accuracy of the avatar redirecting result.

Accordingly, it is possible to first determine whether the initial pose of the reference avatar matches the initial pose of the target avatar. If it is determined that the initial pose of the reference avatar is identical to the initial pose of the target avatar, it is explained that the re-orientation of the target avatar can be accomplished later by migrating a difference between the target pose of the reference avatar and the initial pose of the reference avatar into the target avatar. If it is determined that the two do not match, the initial pose of the target avatar may be adjusted to coincide with the initial pose of the reference avatar. Therefore, the artificial auxiliary workload in the action redirection process of the virtual image is reduced, and the efficiency and the accuracy of the action redirection process can be improved.

In operation S605, a target pose parameter associated with a target pose of a reference avatar is acquired. According to the embodiment of the present disclosure, the manner of acquiring the target pose parameter is similar to that of acquiring the target initial pose parameter for the target avatar, and will not be described again.

In operation S606, an attitude adjustment parameter is obtained according to the target attitude parameter and the reference initial attitude parameter.

In operation S607, a target avatar based on the target pose parameter is generated according to the pose adjustment parameter and the target initial pose parameter of the target avatar.

According to the embodiment of the present disclosure, if the target initial pose parameter needs to be adjusted based on the reference initial pose parameter, the pose adjustment parameter may be obtained based on the target pose parameter and the reference initial pose parameter according to the above-described manner, so that the target avatar based on the target pose parameter is generated using the pose adjustment parameter and the adjusted target initial pose parameter of the target avatar. If the target initial pose parameter does not need to be adjusted, the target avatar based on the target pose parameter can be generated by using the pose adjustment parameter and the target initial pose parameter of the target avatar.

In operation S608, the target avatar is driven according to the target pose parameter to perform an action corresponding to the target pose.

According to the embodiments of the present disclosure, after generating the target avatar based on the target pose parameter, the target avatar may be driven according to the target pose parameter to perform an action corresponding to the target pose. For example, when the target pose represented by the target pose parameters is a wishbone motion, the target avatar may be driven according to the target pose parameters to perform a motion corresponding to wishbone.

According to an embodiment of the present disclosure, a target initial pose parameter of a target avatar is adjusted based on a reference initial pose parameter for a reference avatar such that an initial pose of the adjusted target avatar matches an initial pose of the reference avatar. And then, processing the target initial posture parameters of the adjusted target virtual image based on the posture difference between the target posture and the initial posture of the reference virtual image to obtain a target virtual image based on the target posture parameters, so as to realize the migration of the target posture of the reference virtual image to the target virtual image, namely, the action redirection processing of the target virtual image is completed. The scheme disclosed by the invention is beneficial to reducing the manual auxiliary workload in the action redirection process of the virtual image, and can improve the efficiency of the action redirection processing and the accuracy of the action redirection processing.

Fig. 7 is a block diagram of an avatar generating apparatus according to an embodiment of the present disclosure.

As shown in fig. 7, the avatar generating apparatus 700 includes: a first acquisition module 710, an adjustment module 720, a second acquisition module 730, and a generation module 740.

The first acquisition module 710 is for acquiring a reference initial pose parameter for a reference avatar and a target initial pose parameter for a target avatar. The reference initial pose parameter is used for representing the initial pose of the reference avatar, and the target initial pose parameter is used for representing the initial pose of the target avatar.

The adjustment module 720 is configured to adjust the initial pose parameter of the target based on the reference initial pose parameter, so that the initial pose of the adjusted target avatar matches the initial pose of the reference avatar.

The second acquisition module 730 is configured to acquire a target gesture parameter associated with a target gesture of the reference avatar, and obtain a gesture adjustment parameter according to the target gesture parameter and the reference initial gesture parameter.

The first generation module 740 is configured to generate a target avatar based on the target pose parameter according to the pose adjustment parameter and the adjusted target initial pose parameter of the target avatar.

According to an embodiment of the present disclosure, the reference initial pose parameter includes a first initial local pose matrix of a plurality of reference skeletal nodes of the reference avatar, and the target initial pose parameter includes a second initial local pose matrix of a plurality of target skeletal nodes of the target avatar; the adjustment module 720 includes: a first adjusting unit and a second adjusting unit. The first adjusting unit is used for adjusting the second initial local gesture matrix based on the first initial local gesture matrix corresponding to the second initial local gesture matrix aiming at each of the second initial local gesture matrixes of the plurality of target bone nodes to obtain a third initial local gesture matrix of the target bone node; and the second adjusting unit is used for obtaining the target initial posture parameters of the adjusted target virtual image according to the plurality of third initial local posture matrixes.

According to an embodiment of the present disclosure, the target pose parameters include a first target local pose matrix of a plurality of reference skeletal nodes of the reference avatar; the second acquisition module 730 includes: a first determination unit, a second determination unit, and a third determination unit. The first determining unit is used for determining a first posture adjustment coefficient of the reference skeleton node according to the first target local posture matrix and the first initial local posture matrix aiming at each of the first target local posture matrixes of the plurality of reference skeleton nodes; the second determining unit is used for obtaining a first global posture matrix of the reference skeleton node according to the first initial local posture matrix; and the third determining unit is used for determining the attitude adjustment parameters according to the first attitude adjustment coefficient and the first global attitude matrix.

According to an embodiment of the present disclosure, the first generation module 740 includes: the device comprises a first generating unit, a fourth determining unit, a second generating unit and a third generating unit. The first generation unit is used for obtaining a second global gesture matrix of the target skeleton node according to the third initial local gesture matrix of the target skeleton node aiming at each of the plurality of third initial local gesture matrixes; the fourth determining unit is used for determining a second posture adjustment coefficient according to the posture adjustment parameter and the second global posture matrix; the second generation unit is used for obtaining a second target local posture matrix of the target skeleton node according to the second posture adjustment coefficient and the third initial local posture matrix; and the third generating unit is used for obtaining the target virtual image based on the target posture parameters according to the second target local posture matrix.

According to an embodiment of the present disclosure, the apparatus 700 further includes: the device comprises a third acquisition module, a determination module and a second generation module. The third acquisition module is used for acquiring target attitude parameters associated with the target attitude of the reference avatar in response to determining that the initial attitude of the reference avatar is matched with the initial attitude of the target avatar; the determining module is used for obtaining an attitude adjustment parameter according to the target attitude parameter and the reference initial attitude parameter; and the second generation module is used for generating the target virtual image based on the target posture parameter according to the posture adjustment parameter and the target initial posture parameter aiming at the target virtual image.

According to an embodiment of the present disclosure, the second generating module includes: a fourth generation unit, a fifth determination unit, a fifth generation unit, and a sixth generation unit. The fourth generation unit is used for obtaining a third global gesture matrix of the target skeleton node according to each of the second initial local gesture matrixes of the target skeleton nodes; the fifth determining unit is used for determining a third posture adjustment coefficient according to the posture adjustment parameter and the third global posture matrix; the fifth generation unit is used for obtaining a third target local posture matrix of the target skeleton node according to the third posture adjustment coefficient and the second initial local posture matrix; and the sixth generation unit is used for obtaining the target virtual image based on the target posture parameters according to the third target local posture matrix.

According to an embodiment of the present disclosure, the first adjusting unit includes: a first adjustment subunit and a second adjustment subunit. The first adjusting subunit is used for obtaining a first global posture matrix of the reference skeleton node according to the first initial local posture matrix; and the second adjusting subunit is used for adjusting the second initial local gesture matrix according to the first global gesture matrix of the reference skeleton node to obtain a third initial local gesture matrix of the target skeleton node.

According to an embodiment of the present disclosure, the adjustment module includes: an acquisition unit and a third adjustment unit. The acquisition unit is used for acquiring preset gesture parameters corresponding to the preset gestures; and the third adjusting unit is used for respectively adjusting the reference initial posture parameter and the target initial posture parameter of the target virtual image based on the preset posture parameter, so that the initial posture of the adjusted reference virtual image and the initial posture of the target virtual image are respectively matched with the preset posture.

According to an embodiment of the present disclosure, the apparatus 700 further includes: and a driving module. The driving module is used for driving the target virtual image to execute actions corresponding to the target gesture according to the target gesture parameters.

It should be noted that, in the embodiment of the apparatus portion, the implementation manner, the solved technical problem, the realized function, and the achieved technical effect of each module/unit/subunit and the like are the same as or similar to the implementation manner, the solved technical problem, the realized function, and the achieved technical effect of each corresponding step in the embodiment of the method portion, and are not described herein again.

In the technical scheme of the disclosure, the related processes of collecting, storing, using, processing, transmitting, providing, disclosing and the like of the personal information of the user accord with the regulations of related laws and regulations, and the public order colloquial is not violated.

In the technical scheme of the disclosure, the authorization or consent of the user is obtained before the personal information of the user is obtained or acquired.

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

According to an embodiment of the present disclosure, an electronic device includes: 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 to enable the at least one processor to perform a method as in an embodiment of the present disclosure.

According to an embodiment of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform a method as in an embodiment of the present disclosure.

According to an embodiment of the present disclosure, a computer program product comprising a computer program which, when executed by a processor, implements a method as an embodiment of the present disclosure.

Fig. 8 illustrates a schematic block diagram of an example electronic device 800 that may be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 8, the apparatus 800 includes a computing unit 801 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM) 802 or a computer program loaded from a storage unit 808 into a Random Access Memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other by a bus 804. An input/output (I/O) interface 805 is also connected to the bus 804.

Various components in device 800 are connected to I/O interface 805, including: an input unit 806 such as a keyboard, mouse, etc.; an output unit 807 such as various types of displays, speakers, and the like; a storage unit 808, such as a magnetic disk, optical disk, etc.; and a communication unit 809, such as a network card, modem, wireless communication transceiver, or the like. The communication unit 809 allows the device 800 to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.

The computing unit 801 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 801 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized 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 calculation unit 801 performs the respective methods and processes described above, for example, an avatar generation method. For example, in some embodiments, the avatar generation method may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program may be loaded and/or installed onto device 800 via ROM 802 and/or communication unit 809. When a computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the avatar generation method described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to perform the avatar generation method by any other suitable means (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These 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 the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be 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 this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here 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 a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present disclosure may be performed in parallel, sequentially, or in a different order, provided that the desired results of the disclosed aspects are achieved, and are not limited herein.

The above detailed description should not be taken as limiting the scope of the present disclosure. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present disclosure are intended to be included within the scope of the present disclosure.

Claims

1. An avatar generation method, comprising:

acquiring a reference initial posture parameter for a reference avatar and a target initial posture parameter for a target avatar; the reference initial pose parameter is used for representing the initial pose of the reference avatar, the target initial pose parameter is used for representing the initial pose of the target avatar, wherein the reference initial pose parameter comprises the relative position relationship between the child nodes and the father nodes of a plurality of reference skeleton nodes of the reference avatar, and the target initial pose parameter comprises the relative position relationship between the child nodes and the father nodes of a plurality of target skeleton nodes of the target avatar;

Adjusting the initial posture parameter of the target based on the initial posture parameter of the reference, so that the initial posture of the adjusted target virtual image is matched with the initial posture of the reference virtual image;

acquiring a target posture parameter associated with a target posture of a reference virtual image, and acquiring a posture adjustment parameter according to a posture difference between the target posture parameter and the reference initial posture parameter; and

and generating the target virtual image based on the target posture parameters according to the posture adjustment parameters and the target initial posture parameters of the adjusted target virtual image.

2. The method of claim 1, wherein the reference initial pose parameters comprise a first initial local pose matrix of a plurality of reference skeletal nodes of a reference avatar, and the target initial pose parameters comprise a second initial local pose matrix of a plurality of target skeletal nodes of a target avatar;

the adjusting the target initial pose parameter based on the reference initial pose parameter comprises:

for each of a second initial local pose matrix of a plurality of target bone nodes, adjusting the second initial local pose matrix based on a first initial local pose matrix corresponding to the second initial local pose matrix, resulting in a third initial local pose matrix of the target bone node; and

And obtaining target initial posture parameters of the adjusted target virtual image according to the plurality of third initial local posture matrixes.

3. The method of claim 2, wherein the target pose parameters comprise a first target local pose matrix of a plurality of reference skeletal nodes of a reference avatar, the deriving the pose adjustment parameters from the target pose parameters and the reference initial pose parameters comprising:

determining, for each of a first target local pose matrix of a plurality of reference skeletal nodes, a first pose adjustment coefficient for the reference skeletal node according to the first target local pose matrix and the first initial local pose matrix;

obtaining a first global gesture matrix of the reference skeleton node according to the first initial local gesture matrix; and

and determining the attitude adjustment parameters according to the first attitude adjustment coefficient and the first global attitude matrix.

4. The method of claim 3, wherein the generating the target avatar based on the target pose parameters according to the pose adjustment parameters and the adjusted target initial pose parameters of the target avatar comprises:

For each of the plurality of third initial local pose matrices, obtaining a second global pose matrix of the target bone node according to the third initial local pose matrix of the target bone node;

determining a second posture adjustment coefficient according to the posture adjustment parameter and the second global posture matrix;

obtaining a second target local posture matrix of the target skeleton node according to the second posture adjustment coefficient and the third initial local posture matrix; and

and obtaining the target virtual image based on the target attitude parameters according to the second target local attitude matrix.

5. A method according to claim 3, further comprising:

in response to determining that the initial pose of the reference avatar matches the initial pose of the target avatar,

and generating the target virtual image based on the target posture parameters according to the posture adjustment parameters and the target initial posture parameters aiming at the target virtual image.

6. The method of claim 5, wherein the generating the target avatar based on the target pose parameters according to the pose adjustment parameters and the target initial pose parameters for the target avatar comprises:

Aiming at each of second initial local gesture matrixes of a plurality of target skeleton nodes, obtaining a third global gesture matrix of the target skeleton node according to the second initial local gesture matrix of the target skeleton node;

determining a third attitude adjustment coefficient according to the attitude adjustment parameter and the third global attitude matrix;

obtaining a third target local posture matrix of the target skeleton node according to the third posture adjustment coefficient and the second initial local posture matrix; and

and obtaining the target virtual image based on the target attitude parameters according to the third target local attitude matrix.

7. The method of any of claims 2 to 6, wherein the adjusting the second initial local pose matrix based on a first initial local pose matrix corresponding to the second initial local pose matrix, resulting in a third initial local pose matrix of the target bone node comprises:

and adjusting the second initial local gesture matrix according to the first global gesture matrix of the reference skeleton node to obtain a third initial local gesture matrix of the target skeleton node.

8. The method of claim 1, wherein the adjusting the target initial pose parameter based on the reference initial pose parameter comprises:

acquiring preset gesture parameters corresponding to the preset gestures; and

and respectively adjusting the reference initial posture parameter and the target initial posture parameter of the target virtual image based on the preset posture parameter, so that the initial posture of the adjusted reference virtual image and the initial posture of the target virtual image are respectively matched with the preset posture.

9. The method of any one of claims 1 to 6, further comprising:

and driving the target virtual image to execute actions corresponding to the target gesture according to the target gesture parameters.

10. An avatar generation apparatus comprising:

a first acquisition module for acquiring a reference initial pose parameter for a reference avatar and a target initial pose parameter for a target avatar; the reference initial pose parameter is used for representing the initial pose of the reference avatar, the target initial pose parameter is used for representing the initial pose of the target avatar, wherein the reference initial pose parameter comprises the relative position relationship between the child nodes and the father nodes of a plurality of reference skeleton nodes of the reference avatar, and the target initial pose parameter comprises the relative position relationship between the child nodes and the father nodes of a plurality of target skeleton nodes of the target avatar;

The adjusting module is used for adjusting the initial posture parameters of the target based on the initial posture parameters of the reference, so that the initial posture of the adjusted target virtual image is matched with the initial posture of the reference virtual image;

the second acquisition module is used for acquiring target attitude parameters related to the target attitude of the reference virtual image and acquiring attitude adjustment parameters according to the attitude difference between the target attitude parameters and the reference initial attitude parameters; and

and the first generation module is used for generating the target virtual image based on the target posture parameter according to the posture adjustment parameter and the adjusted target initial posture parameter of the target virtual image.

11. The apparatus of claim 10, wherein the reference initial pose parameters comprise a first initial local pose matrix of a plurality of reference skeletal nodes of a reference avatar, and the target initial pose parameters comprise a second initial local pose matrix of a plurality of target skeletal nodes of a target avatar; the adjustment module includes:

a first adjustment unit, configured to adjust, for each of second initial local pose matrices of a plurality of target bone nodes, the second initial local pose matrix based on a first initial local pose matrix corresponding to the second initial local pose matrix, resulting in a third initial local pose matrix of the target bone node; and

And the second adjusting unit is used for obtaining the target initial posture parameters of the adjusted target virtual image according to the plurality of third initial local posture matrixes.

12. The apparatus of claim 11, wherein the target pose parameters comprise a first target local pose matrix of a plurality of reference skeletal nodes of a reference avatar; the second acquisition module includes:

a first determining unit, configured to determine, for each of a first target local pose matrix of a plurality of reference skeletal nodes, a first pose adjustment coefficient of the reference skeletal node according to the first target local pose matrix and the first initial local pose matrix;

the second determining unit is used for obtaining a first global gesture matrix of the reference skeleton node according to the first initial local gesture matrix; and

and the third determining unit is used for determining the attitude adjustment parameters according to the first attitude adjustment coefficient and the first global attitude matrix.

13. The apparatus of claim 12, wherein the first generation module comprises:

a first generating unit, configured to obtain, for each of a plurality of third initial local pose matrices, a second global pose matrix of the target skeletal node according to the third initial local pose matrix of the target skeletal node;

A fourth determining unit, configured to determine a second posture adjustment coefficient according to the posture adjustment parameter and the second global posture matrix;

the second generating unit is used for obtaining a second target local posture matrix of the target skeleton node according to the second posture adjustment coefficient and the third initial local posture matrix; and

and the third generation unit is used for obtaining the target virtual image based on the target posture parameters according to the second target local posture matrix.

14. The apparatus of claim 12, further comprising:

and a second generation module for generating a target avatar based on the target pose parameter according to the pose adjustment parameter and the target initial pose parameter for the target avatar in response to determining that the initial pose of the reference avatar matches the initial pose of the target avatar.

15. The apparatus of claim 14, wherein the second generation module comprises:

a fourth generating unit, configured to obtain, for each of second initial local pose matrices of a plurality of target skeletal nodes, a third global pose matrix of the target skeletal node according to the second initial local pose matrix of the target skeletal node;

A fifth determining unit, configured to determine a third posture adjustment coefficient according to the posture adjustment parameter and the third global posture matrix;

a fifth generating unit, configured to obtain a third target local pose matrix of the target skeletal node according to the third pose adjustment coefficient and the second initial local pose matrix; and

and a sixth generating unit, configured to obtain the target avatar based on the target pose parameter according to the third target local pose matrix.

16. The apparatus according to any one of claims 11 to 15, wherein the first adjustment unit comprises:

the first adjusting subunit is used for obtaining a first global gesture matrix of the reference skeleton node according to the first initial local gesture matrix; and

and the second adjusting subunit is used for adjusting the second initial local gesture matrix according to the first global gesture matrix of the reference skeleton node to obtain a third initial local gesture matrix of the target skeleton node.

17. The apparatus of claim 10, wherein the adjustment module comprises:

the acquisition unit is used for acquiring preset gesture parameters corresponding to the preset gestures; and

And the third adjusting unit is used for respectively adjusting the reference initial posture parameter and the target initial posture parameter of the target virtual image based on the preset posture parameter so that the initial posture of the adjusted reference virtual image and the initial posture of the target virtual image are respectively matched with the preset posture.

18. The apparatus of any of claims 10 to 15, further comprising:

and the driving module is used for driving the target virtual image to execute actions corresponding to the target gesture according to the target gesture parameters.

19. 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 to enable the at least one processor to perform the method of any one of claims 1 to 9.

20. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1 to 9.