CN113192164A - Avatar follow-up control method and device, electronic equipment and readable storage medium - Google Patents

Avatar follow-up control method and device, electronic equipment and readable storage medium Download PDF

Info

Publication number
CN113192164A
CN113192164A CN202110517578.4A CN202110517578A CN113192164A CN 113192164 A CN113192164 A CN 113192164A CN 202110517578 A CN202110517578 A CN 202110517578A CN 113192164 A CN113192164 A CN 113192164A
Authority
CN
China
Prior art keywords
template
expression base
features
follow
templates
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110517578.4A
Other languages
Chinese (zh)
Inventor
卫华威
韩欣彤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Huya Technology Co Ltd
Original Assignee
Guangzhou Huya Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangzhou Huya Technology Co Ltd filed Critical Guangzhou Huya Technology Co Ltd
Priority to CN202110517578.4A priority Critical patent/CN113192164A/en
Publication of CN113192164A publication Critical patent/CN113192164A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T13/00—Animation
    • G06T13/20—3D [Three Dimensional] animation
    • G06T13/40—3D [Three Dimensional] animation of characters, e.g. humans, animals or virtual beings
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/21—Server components or server architectures
    • H04N21/218—Source of audio or video content, e.g. local disk arrays
    • H04N21/2187—Live feed
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
    • H04N21/258—Client or end-user data management, e.g. managing client capabilities, user preferences or demographics, processing of multiple end-users preferences to derive collaborative data
    • H04N21/25866—Management of end-user data
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
    • H04N21/262—Content or additional data distribution scheduling, e.g. sending additional data at off-peak times, updating software modules, calculating the carousel transmission frequency, delaying a video stream transmission, generating play-lists
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
    • H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
    • H04N21/44213—Monitoring of end-user related data
    • H04N21/44218—Detecting physical presence or behaviour of the user, e.g. using sensors to detect if the user is leaving the room or changes his face expression during a TV program
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
    • H04N21/4508—Management of client data or end-user data

Abstract

The application provides a virtual image follow-up control method, a virtual image follow-up control device, an electronic device and a readable storage medium, wherein a plurality of face features contained in a face image are obtained by obtaining a face image of a anchor, a plurality of preset expression base templates corresponding to each face feature can be obtained according to each face feature, and then template coefficients of the face features relative to each expression base template are determined according to the face features and the plurality of expression base templates. And performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image based on each expression base template and the corresponding template coefficients. According to the scheme, the facial features of the anchor are embodied by the expression base template carrying the template coefficients, so that the virtual image is controlled, follow-up control over the virtual image can be achieved under the condition that other auxiliary equipment is not needed, and the control information is more standard and more accurate in control by controlling the information of the expression base template.

Description

Avatar follow-up control method and device, electronic equipment and readable storage medium
Technical Field
The application relates to the technical field of live broadcast, in particular to a follow-up control method and device for an avatar, electronic equipment and a readable storage medium.
Background
With the popularity of the live broadcast industry, more and more people want to enter the live broadcast industry, but part of anchor broadcasters cannot self-trust the appearance of the anchor broadcasters, and if the anchor broadcasters can be assisted by virtual images, the virtual images can perform vivid performance along with the anchor broadcasters, the threshold of the live broadcast is extremely reduced.
In the existing schemes, some anchor casts directly use the head portrait stickers or live broadcast with head covers, but the static stickers or the head covers cannot follow the anchor casts to make vivid expressions, which results in great discount on the interaction effect with audiences. Another solution is to let the anchor wear a professional kinetic catching device, but this approach is costly and causes great discomfort to the anchor by the anchor wearing the helmet all the time.
Disclosure of Invention
The object of the present application includes, for example, providing an avatar follow-up control method, apparatus, electronic device, and readable storage medium, which can implement the follow-up control of an avatar accurately, normatively, and without interference.
The embodiment of the application can be realized as follows:
in a first aspect, the present application provides an avatar following control method, the method comprising:
acquiring a human face image of a anchor, and acquiring a plurality of human face features contained in the human face image;
aiming at each face feature, obtaining a plurality of preset expression base templates corresponding to the face feature;
determining template coefficients of the facial features relative to each expression base template according to the facial features and the expression base templates;
and performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image based on each expression base template and the corresponding template coefficients.
In an alternative embodiment, the step of determining, according to the facial feature and the expression base templates, a template coefficient of the facial feature with respect to each of the expression base templates includes:
and optimizing the objective function according to a pre-established objective function, the facial features and the expression base templates until a plurality of template coefficients contained in the objective function are obtained when a preset requirement is met, wherein the template coefficients correspond to the expression base templates one to one.
In an optional embodiment, the step of performing optimization processing on the objective function according to a pre-established objective function, the facial features, and the expression base templates includes:
and performing projection transformation on each expression base template, and performing optimization processing on the objective function according to a pre-established objective function, the human face characteristics and the plurality of expression base templates after projection transformation.
In an optional embodiment, the step of performing projection transformation on each expression base template, and performing optimization processing on the objective function according to a pre-established objective function, the facial features, and a plurality of expression base templates after projection transformation includes:
performing primary projection transformation on each expression base template by using a preset initial transformation coefficient, and obtaining a plurality of template coefficients contained in the objective function according to a pre-established objective function, the human face features and a plurality of expression base templates subjected to primary projection transformation;
and taking the obtained template coefficients as the updated transformation coefficients of the expression base templates, performing re-projection transformation on the expression base templates by using the updated transformation coefficients, and performing optimization processing on the objective function according to the objective function, the human face features and the plurality of expression base templates after the re-projection transformation until the preset requirements are met.
In an alternative embodiment, the facial features comprise a plurality of facial feature points, and each expression base template comprises a plurality of expression feature points;
the step of performing projection transformation on each expression base template comprises the following steps:
determining expression feature points corresponding to the facial feature points in each expression base template aiming at each facial feature point in the facial features;
and carrying out projection transformation on each expression base template according to the coordinate values of the expression feature points in each expression base template and the coordinate values of the corresponding face feature points.
In an alternative embodiment, the facial features comprise a plurality of facial feature points, and each expression base template comprises a plurality of expression feature points;
the step of optimizing the objective function according to the pre-established objective function, the facial features and the expression base templates after projection transformation comprises the following steps:
obtaining a first array constructed by the plurality of human face feature points, and obtaining a second array constructed by a plurality of expression feature points contained in each expression base template;
and based on a pre-established target function, performing weighted accumulation on the plurality of second arrays according to corresponding template coefficients, subtracting the first array, and determining each template coefficient when the target function meets the preset requirement.
In an optional embodiment, the step of performing state following control on virtual features corresponding to the face features in the constructed virtual image based on each expression base template and the corresponding template coefficients includes:
carrying out weighted accumulation on the plurality of expression base templates according to corresponding template coefficients to obtain control information;
and performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image according to the control information.
In an optional embodiment, the control information includes a plurality of pieces of coordinate information, and each piece of coordinate information is a weighted result of coordinates of corresponding expression feature points included in a plurality of expression base templates;
the step of performing state follow-up control on virtual features corresponding to the face features in the constructed virtual image according to the control information comprises the following steps:
acquiring virtual features corresponding to the human face features in the constructed virtual image, and acquiring a plurality of virtual feature points contained in the virtual features;
and controlling each virtual feature point in the virtual features according to the corresponding coordinate information in the control information so as to drive the virtual image to carry out state follow-up.
In a second aspect, the present application provides an avatar following control apparatus, the apparatus comprising:
the first acquisition module is used for acquiring a human face image of a anchor and acquiring a plurality of human face features contained in the human face image;
the second obtaining module is used for obtaining a plurality of corresponding expression base templates aiming at each face feature;
the determining module is used for determining a template coefficient of the face feature relative to each expression base template according to the face feature and the expression base templates;
and the control module is used for carrying out state follow-up control on virtual features corresponding to the human face features in the constructed virtual image based on each expression base template and the corresponding template coefficient.
In a third aspect, the present application provides an electronic device comprising one or more storage media and one or more processors in communication with the storage media, the one or more storage media storing processor-executable machine-executable instructions that, when executed by the electronic device, are executed by the processors to perform the method steps of any one of the preceding embodiments.
In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon machine-executable instructions which, when executed, implement the method steps of any one of the preceding embodiments.
The beneficial effects of the embodiment of the application include, for example:
the application provides a virtual image follow-up control method, a virtual image follow-up control device, an electronic device and a readable storage medium, wherein a plurality of face features contained in a face image are obtained by obtaining a face image of a anchor, a plurality of preset expression base templates corresponding to each face feature can be obtained according to each face feature, and then template coefficients of the face features relative to each expression base template are determined according to the face features and the plurality of expression base templates. And performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image based on each expression base template and the corresponding template coefficients. According to the scheme, the facial features of the anchor are embodied by the expression base template carrying the template coefficients, so that the virtual image is controlled, follow-up control over the virtual image can be achieved under the condition that other auxiliary equipment is not needed, and the control information is more standard and more accurate in control by controlling the information of the expression base template.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 is a schematic view of an application scenario of an avatar servo control method according to an embodiment of the present application;
fig. 2 is a flowchart of an avatar follow-up control method according to an embodiment of the present application;
fig. 3 is a schematic diagram of a face template provided in an embodiment of the present application;
FIG. 4 is a schematic diagram of a face feature point provided in an embodiment of the present application;
FIG. 5 is a schematic diagram illustrating feature point association between an expression base template and facial features in an embodiment of the present application;
FIG. 6 is a flowchart of sub-steps included in step S130 of FIG. 2;
FIG. 7 is a flowchart of sub-steps included in step S140 of FIG. 2;
FIG. 8 is a schematic view of an avatar provided in an embodiment of the present application;
fig. 9 is a block diagram of an electronic device according to an embodiment of the present application;
fig. 10 is a functional block diagram of an avatar servo control apparatus according to an embodiment of the present application.
Icon: 100-live broadcast providing terminal; 200-a live broadcast server; 300-a live broadcast receiving end; 410-a memory; 420-a processor; 430-avatar follow-up control; 431-a first obtaining module; 432-a second obtaining module; 433-a determination module; 434 — control module.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
In the description of the present application, it is noted that the terms "first", "second", and the like are used merely for distinguishing between descriptions and are not intended to indicate or imply relative importance.
It should be noted that the features of the embodiments of the present application may be combined with each other without conflict.
As shown in fig. 1, an application scene schematic diagram of the avatar follow-up control method provided in the embodiment of the present application may include a live broadcast providing end 100, a live broadcast receiving end 300, and a live broadcast server 200 in communication connection with the live broadcast providing end 100 and the live broadcast receiving end 300, respectively.
The live broadcast providing terminal 100 may be a terminal device (such as a mobile phone, a tablet computer, a computer, etc.) used by a main broadcast during live broadcast, and the live broadcast receiving terminal 300 may be a terminal device (such as a mobile phone, a tablet computer, a computer, etc.) used by an audience during live broadcast watching.
In this embodiment, a video capture device for capturing video frames of the anchor may be further included in the scene, and the video capture device may be, but is not limited to, a camera, a lens of a digital camera, a monitoring camera, a webcam, or the like.
The video capture device may be directly installed or integrated in the live broadcast provider 100. For example, the video capture device may be a camera configured on the live provider 100, and other modules or components in the live provider 100 may receive videos, images, and the like sent from the video capture device via the internal bus. Alternatively, the video capture device may be independent of the live broadcast provider 100, and the two devices may communicate with each other in a wired or wireless manner.
The live video provider 100 may send the live video stream to the live server 200, and the viewer may access the live server 200 through the live receiver 300 to watch the live video.
With reference to fig. 2, an embodiment of the present application further provides an avatar follow-up control method applicable to an electronic device, for performing follow-up control on an avatar in a live video. The electronic device may be the live broadcast provider 100 or the live broadcast server 200. The method steps defined by the process related to the avatar animation control method may be implemented by the electronic device. The specific process shown in fig. 2 will be described in detail below.
Step S110, a human face image of a main broadcaster is obtained, and a plurality of human face features contained in the human face image are obtained.
And step S120, aiming at each facial feature, obtaining a plurality of preset expression base templates corresponding to the facial features.
Step S130, determining a template coefficient of the facial features relative to each expression base template according to the facial features and the expression base templates.
And step S140, performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image based on each expression base template and the corresponding template coefficients.
When the anchor performs live broadcasting by using the live broadcasting provider 100, the video capture device can capture live video frames of the anchor. The live broadcast provider 100 may send the captured live broadcast video frames to the live broadcast server 200, and push the live broadcast video frames through the live broadcast server 200 to the live broadcast receiver 300 held by each viewer watching the anchor.
In the live broadcasting process of the anchor, the face information of the anchor is information mainly representing the state of the anchor, and in practical application, the anchor may need to utilize an avatar, such as a cartoon character, to synchronously display or replace the face state of the anchor by controlling the avatar, so as to carry out live broadcasting by means of the avatar.
In the process, how to make the virtual image accurately keep synchronous follow-up with the real-time state of the anchor is particularly important.
In this embodiment, the live broadcast providing terminal 100 may collect a real-time video frame of a main broadcast, the main broadcast providing terminal may perform a subsequent processing process based on the collected video frame, and may also send the collected video frame to the live broadcast server 200, and the live broadcast server 200 performs the subsequent processing process based on the received video frame, which is not limited in this embodiment.
And performing face recognition processing on the acquired video frame to obtain a face image contained in the video frame. The process may include processes of recognizing a face region in a video frame, extracting the face region to form a face image, and the like.
The face image includes a plurality of face features, such as left eye, right eye, nose, mouth, eyebrow, etc. When the virtual image is controlled to follow the anchor state, follow-up calculation can be performed respectively aiming at each human face characteristic.
In this embodiment, a plurality of preset face templates are also prestored in the electronic device, and optionally, a general open source model may be used as the face template. Each face template includes a plurality of expression base templates, for example, ten different face templates as shown in fig. 3. The expression base template such as the first face template comprises a left eye template, a right eye template, a mouth template, a nose template and the like.
It should be noted that each face template includes expression base templates of the same type, but the corresponding expression base templates in different face templates may be in different states, for example, a left eye template in some face templates is in an eye-open state, and a left eye template in some face templates is in an eye-closed state.
For each obtained facial feature of the anchor, a plurality of preset expression base templates corresponding to the facial feature may be obtained, for example, for a left-eye feature of the anchor, a left-eye template in each facial template may be obtained, and the obtained plurality of left-eye templates may be in different states, such as an eye-open state, an eye-closed state, a half-open eye state, and the like.
By analyzing the obtained face features and the information of the plurality of expression base templates corresponding to the face features, the template coefficients of the face features relative to the expression base templates can be determined. The value of each template coefficient is in the range (0,1), that is, the probability of each template coefficient representing the corresponding expression base template to represent the face features is high.
Through the calculation, the facial features of the anchor can be converted into a plurality of expression base templates to be embodied, specifically, a plurality of expression base templates carrying respective template coefficients.
On the basis, the electronic equipment also comprises pre-constructed virtual images, such as cartoon images, romantic images, hero images and the like, and particularly, the corresponding setting can be carried out according to the requirements.
Similarly, the avatar includes a plurality of virtual features, which may include left eye, right eye, nose, mouth, etc. The human face features of the anchor correspond to the virtual features of the virtual image one by one, and for example, the left eye of the virtual image can be subjected to follow-up control based on the feature information of the left eye of the anchor. Of course, in the actual process, a mode of synchronously analyzing and processing each face feature of the anchor in parallel and finally synchronously controlling each virtual feature of the virtual image is generally adopted, that is, the full-face feature of the virtual image is controlled based on the full-face feature of the anchor. Only when the scheme is explained in the embodiment, the processing of each face feature is explained.
After the plurality of expression base templates corresponding to a certain face feature of the anchor and the template coefficients corresponding to the expression base templates are obtained through the process, when the virtual image is controlled, the corresponding virtual feature in the virtual image can be subjected to state follow-up control based on the obtained expression base templates and the template coefficients corresponding to the expression base templates.
Because the facial features of the anchor are converted into the corresponding expression base templates carrying the template coefficients to be embodied, the virtual image is controlled based on the expression base templates carrying the template coefficients, and the effect that the virtual image follows the face state of the anchor can be achieved. In addition, the expression base template is a template with the preset information specification, so that the control information is more standard and more accurate in control by controlling the preset template information in practical application.
In this embodiment, the facial features of the anchor are converted into information of an expression base template carrying template coefficients to be embodied, where the template coefficients represent how high probability the expression base template embodies the facial features, and the accurate determination of the template coefficients is particularly important for whether the expression base template can accurately embody the facial features. Because, in the present embodiment, in order to accurately determine the template coefficients, the template coefficients are determined in the following manner:
and optimizing the objective function according to the pre-established objective function, the human face characteristics and the plurality of expression base templates until a plurality of template coefficients contained in the objective function are obtained when the preset requirements are met.
And the template coefficients correspond to the expression base templates one by one.
In this embodiment, an objective function may be pre-constructed, where the objective function includes two components, the first component is expression base template information and template coefficients, and the second component is face feature information. Under the condition that the expression base template and the facial features are known, the information of the expression base template and the facial features is brought into an objective function, and the optimization of the objective function is substantially the optimization of template coefficients contained in the objective function.
In an implementation manner, in the optimization process of the objective function, when a preset requirement is met, it may be determined that the template coefficient at this time meets the preset requirement, and the template coefficient may be fixed. The mode of judging that the preset requirement is met may be that the iteration number reaches the preset number, or that the function value of the objective function is lower than a preset threshold value, and the like, and may be specifically set according to the requirement.
In this embodiment, a face template pre-constructed in an electronic device is a 3D model, and accordingly, information of an expression base template in the face template is 3D information, and a face image obtained based on a video frame is a two-dimensional image, so that to accurately perform optimization of an objective function based on the expression base template and a face feature, it is first necessary to keep the 3D model information consistent with the two-dimensional information of the face feature, and therefore, in this embodiment, the process of performing optimization processing on the objective function may be implemented in the following manner:
and performing projection transformation on each expression base template, and performing optimization processing on the target function according to a pre-established target function, the human face characteristics and the plurality of expression base templates after projection transformation.
In this embodiment, it is considered that in the actual application process, the face of the anchor may not be facing the video capture device in many times in the live broadcast process, that is, in the obtained face image of the anchor, the face region may not be a front face image. Each face template pre-established in the electronic device is generally established based on a front face image, that is, each expression base template is also a feature of the front face in the view angle.
Therefore, when the expression base template is subjected to projection transformation, in addition to projecting the expression base template from the 3D model to the two-dimensional coordinate system to keep the coordinate system information consistent, in order to align the expression base template with the human face features of the anchor as much as possible, in this embodiment, the projected expression base template may be subjected to processes such as shifting and scaling after being projected, so that the difference between the projected expression base template and the human face features is reduced as much as possible.
In this embodiment, the anchor face image includes a plurality of face features, and each face feature includes a plurality of face feature points, for example, for the anchor left eye, the left eye may be composed of a plurality of eye feature points, such as a plurality of feature points constituting an eyebox contour, a center point for locating an eyebox position, and the like.
Similarly, each face template comprises a plurality of expression base templates, each expression base template comprises a plurality of expression feature points, and similarly, for a left eye template in the expression base templates, each left eye template is composed of a plurality of feature points. In addition, in order to facilitate management of each expression base template belonging to the same type (such as left eye) but being in different states (such as open eye, closed eye, half open eye, and the like), feature points of the expression base templates of the same type are encoded in the same encoding manner, for example, corresponding expression feature points in each expression base template are marked by the same serial number. That is, points, edges and connection modes of points representing feature points in each expression base template are all consistent, and points with a certain serial number represent the same semantic meaning, for example, the nth point in each expression base template is fixedly represented as an eye corner point.
Therefore, in the process of performing projection transformation on each expression base template to the facial features, first, for each facial feature of the anchor, a plurality of facial feature points constituting the facial feature may be extracted, where the plurality of facial feature points are 2D feature points, for example, as shown in fig. 4. It should be noted that the identification manner of the face feature points may be consistent with the identification manner of the template feature points in the expression base template.
On this basis, for each facial feature point in the facial features, the expression base feature points corresponding to the facial feature points in each expression base template can be determined, that is, the facial feature points and the corresponding expression base feature points are bound one by one. As shown in fig. 5, the left side is the face feature points in the face feature, and the right side is the 3D template feature points in the face template.
And then, carrying out projection transformation on each expression base template according to the coordinate values of the expression characteristic points in each expression base template and the coordinate values of the corresponding face characteristic points.
That is, each individual face feature point in the face features is extracted, then the face feature points are bound with the corresponding template feature points, and projection transformation of the expression base template is realized based on the binding relationship, the coordinate values of the template feature points and the coordinate values of the face feature points. Therefore, projection transformation is realized based on the coordinate values of the corresponding feature points, and the difference between the expression base template and the human face features after projection transformation can be reduced as much as possible.
In this embodiment, in the process of performing projection transformation on the expression base templates, corresponding transformation coefficients need to be adopted for each expression base template to perform transformation. The transformation results are correspondingly different under different transformation coefficients, and the transformation results further influence the optimization process of the objective function, that is, the projection transformation of the expression base template and the optimization of the objective function are mutually influenced. In view of this, in the present embodiment, the projection transformation and the respective optimization results of the objective function are utilized to further optimize each other, so that the final result satisfies the preset requirement.
In detail, in this embodiment, when performing optimization processing on the objective function, first, for projection transformation of the expression base templates, initial projection transformation may be performed on each expression base template by using a preset initial transformation coefficient, and a plurality of template coefficients included in the objective function are obtained according to a pre-established objective function, a human face feature, and a plurality of expression base templates after the initial projection transformation.
In this embodiment, when the projective transformation of the expression base template is performed for the first time, for example, for a plurality of expression base templates in the standard state, the eye-closed state, the eye-blinking state, and the like, the initial transformation coefficient may be a coefficient group such as (1,0,0, …). And performing first projective transformation by using the initial transformation coefficient.
After the projection transformation, the expression base template subjected to the projection transformation can be brought into the objective function, and the template coefficient of the wheel is calculated.
On the basis of calculating the template coefficients, the obtained template coefficients are used as the updated transformation coefficients of the expression base templates, the updated transformation coefficients are used for carrying out re-projection transformation on the expression base templates, and optimization processing is carried out on the polar target functions of the expression base templates after the re-projection transformation according to the target functions, the human face features and the multiple expression base templates after the re-projection transformation until the preset requirements are met.
Namely, the expression base template is subjected to projection transformation by utilizing the template coefficient calculated in the previous round, and the expression base template subjected to projection transformation is brought into the objective function to calculate the template coefficient of a new round. In this way, after multiple rounds of iterative computations are performed, when the iteration number reaches a preset number, or the function value of the objective function is lower than a preset threshold, it can be determined that the preset requirement is met, and the template coefficient in the objective function at this time can be obtained and used as the final template coefficient.
In this embodiment, the obtained template coefficients may be further optimized by the optimization method under the influence of the projection transformation and the objective function.
In this embodiment, as can be seen from the above description, the objective function is composed of template coefficients, expression base templates, and facial features, where the facial features include a plurality of facial feature points, and each expression base template includes a plurality of template feature points. In detail, referring to fig. 6, the objective function is optimized to determine the template coefficients as follows:
step S131, a first array constructed by the plurality of facial feature points is obtained, and a second array constructed by the plurality of expression feature points included in each expression base template is obtained.
And step S132, based on a pre-established target function, performing weighted accumulation on the plurality of second arrays according to corresponding template coefficients, subtracting the first array, and determining each template coefficient when the target function meets a preset requirement.
In this embodiment, the plurality of facial feature points may form a first array, for example, if the number of extracted facial feature points is 100, a two-dimensional first array of 100 × 2 may be obtained by combining the coordinate values of the facial feature points. Accordingly, each emoji template may correspond to a two-dimensional second array of 100 x 2, since the emoji feature points are defined in a similar manner to the facial feature points.
As one implementation, the pre-constructed objective function may be as follows:
f＝||P(V)-L||2
wherein, V ═ template coefficient 1 ═ expression base template 1+ template coefficient 2 ═ expression base template 2+.
In the above formula, p (v) represents the result of weighted accumulation of each expression base template (each second array) after projective transformation according to the template coefficient, and L represents the first array formed by the human face feature points. And determining each template coefficient through fitting the objective function.
After the facial features are embodied by the expression base template in the above manner, the constructed virtual image can be subjected to follow-up control further based on the determined expression base template and the template coefficient. In detail, referring to fig. 7, this step can be implemented as follows:
and step S141, performing weighted accumulation on the plurality of expression base templates according to corresponding template coefficients to obtain control information.
And S142, performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image according to the control information.
In this embodiment, the obtained plurality of expression base templates may be subjected to weighted accumulation according to corresponding template coefficients to be used as control information. For example, the control information may be an accumulation result calculated according to the following formula:
template coefficient 1 expression basis template 1+ template coefficient 2 expression basis template 2+.
As can be seen from the above, each expression base template includes a plurality of template feature points, and the virtual features in the avatar also include a plurality of virtual feature points, and the control of the avatar is substantially the control of each virtual feature point.
That is, the obtained control information includes a plurality of pieces of coordinate information, and each piece of coordinate information is a weighting result of coordinates of corresponding expression feature points included in the plurality of expression base templates. For example, if the left eye in the face feature is followed, the obtained coordinate information is the coordinate information of the corresponding feature point in the plurality of left eye templates.
When the virtual image is controlled, the virtual feature corresponding to the human face feature in the constructed virtual image can be obtained, and a plurality of virtual feature points contained in the virtual feature can be obtained. For example, the left eye feature of the avatar, a plurality of feature points contained in the left eye of the avatar.
And then each virtual feature point in the virtual features is controlled according to the corresponding coordinate information in the control information so as to drive the virtual image to carry out state follow-up. That is, the coordinates of the corresponding virtual feature points in the virtual features are controlled by using the coordinate information of the corresponding template feature points in the expression base template after weighting according to the template coefficients.
For example, as shown in fig. 8, when the anchor mouth is in a certain state, the state may be represented by a plurality of expression base templates (mouth templates) together with different template coefficients, for example, a template in the open-mouth state and a template in the beep-mouth state, and the state may be characterized by an expression base template of 0.2 × open mouth + an expression base template of 0.5 × beep mouth, for example. Wherein 0.2 is the template coefficient of the expression base template in the open mouth state, and 0.5 is the template coefficient of the expression base template in the Duzui state.
Based on the obtained result, the virtual image can be correspondingly controlled to follow the mouth state of the anchor, so that mouth state follow-up is realized.
By the mode, the human face features of the anchor can be converted into the expression base template carrying the template coefficients to be embodied, and the virtual image is controlled by weighting according to the template coefficients based on the expression base template, so that the purpose that the virtual image follows the human face features of the anchor to perform state follow-up control is achieved.
In an implementation manner, if the electronic device is the live broadcast providing terminal 100, the live broadcast providing terminal 100 may generate a corresponding live broadcast video stream and send the live broadcast video stream to the live broadcast server 200 after acquiring a video frame of a main broadcast and correspondingly controlling an avatar, and the live broadcast video stream is pushed by the live broadcast server 200 to each live broadcast receiving terminal 300 for viewing by a viewer.
In another implementation manner, if the electronic device is the live broadcast server 200 described above, the live broadcast server 200 may generate a push stream from video frames including the anchor and the avatar after implementing the follow-up control of the avatar, and push the push stream to each live broadcast receiving end 300 for the audience to watch.
The virtual image follow-up control method provided by the embodiment of the application can be embodied by converting the human face features of the anchor into the expression base template with the preset standard, and then the virtual image is subjected to follow-up control based on the corresponding expression base template. Under the condition that auxiliary equipment is not needed, the purpose that the virtual image follows the anchor state to carry out follow-up control can be successfully achieved, and the expression base template converted into the preset standard is used for controlling, so that control information is more standard and control is more accurate.
Referring to fig. 9, an electronic device is further provided in an embodiment of the present application. The electronic device may be a live device, for example, a terminal device used by a main broadcast during live broadcast (such as the live broadcast provider 100 described above), or a server communicatively connected to the main broadcast during live broadcast (such as the live broadcast server 200 described above).
In detail, the electronic device may include a memory 410, a processor 420, and an avatar servo control 430. The memory 410 and the processor 420 are electrically connected directly or indirectly to enable data transfer or interaction. For example, they may be electrically connected to each other via one or more communication buses or signal lines. The avatar follow-up control 430 includes at least one software function module that can be stored in the memory 410 in the form of software or firmware (firmware). The processor 420 is configured to execute an executable computer program stored in the memory 410, for example, a software function module and a computer program included in the avatar follow-up control device 430, so as to implement an avatar follow-up control method, thereby ensuring that the avatar is controlled to follow a main broadcasting state for live broadcasting display, and ensuring that the follow-up effect of the avatar is more standard and accurate.
The Memory 410 may be, but is not limited to, a Random Access Memory 410 (RAM), a Read Only Memory (ROM), a Programmable Read Only Memory (PROM), an Erasable Read Only Memory (EPROM), an electrically Erasable Read Only Memory (EEPROM), and the like. The memory 410 is used for storing programs, and the processor 420 executes the programs after receiving execution instructions.
The processor 420 may be an integrated circuit chip having signal processing capabilities. For example, the System may be a Central Processing Unit (CPU), a Network Processor (NP), a System on Chip (SoC), a Digital Signal Processor (DSP), etc. to implement or execute the methods and steps disclosed in the embodiments of the present application.
It will be appreciated that the configuration shown in fig. 9 is merely illustrative, and that the electronic device may also include more or fewer components than shown in fig. 9, or have a different configuration than shown in fig. 9, for example, and may also include a communication unit for information interaction with other live devices. Wherein the components shown in fig. 9 may be implemented in hardware, software, or a combination thereof.
Please refer to fig. 10, which is a block diagram of functional modules of an avatar follow-up control apparatus 430 according to an embodiment of the present application. The avatar follow-up control 430 includes a first obtaining module 431, a second obtaining module 432, a determining module 433, and a control module 434.
The first obtaining module 431 is configured to obtain a face image of a main broadcaster, and obtain a plurality of face features included in the face image.
In this embodiment, the first obtaining module 431 may be configured to perform step S110 shown in fig. 2, and reference may be made to the foregoing description of step S110 for relevant contents of the first obtaining module 431.
A second obtaining module 432, configured to obtain, for each facial feature, a plurality of corresponding expression base templates.
In this embodiment, the second obtaining module 432 may be configured to perform step S120 shown in fig. 2, and reference may be made to the foregoing description of step S120 for relevant contents of the second obtaining module 432.
A determining module 433, configured to determine, according to the facial feature and the expression base templates, a template coefficient of the facial feature with respect to each expression base template.
In this embodiment, the determining module 433 may be configured to perform step S130 shown in fig. 2, and reference may be made to the foregoing description of step S130 for relevant contents of the determining module 433.
And a control module 434, configured to perform state following control on virtual features corresponding to the face features in the constructed virtual image based on each expression base template and the corresponding template coefficients.
In this embodiment, the control module 434 may be configured to execute step S140 shown in fig. 2, and reference may be made to the foregoing description of step S140 for relevant contents of the control module 434.
In a possible implementation manner, the determining module 433 may be specifically configured to:
and optimizing the objective function according to a pre-established objective function, the facial features and the expression base templates until a plurality of template coefficients contained in the objective function are obtained when a preset requirement is met, wherein the template coefficients correspond to the expression base templates one to one.
In a possible implementation manner, the determining module 433 may be configured to perform an optimization process by:
and performing projection transformation on each expression base template, and performing optimization processing on the objective function according to a pre-established objective function, the human face characteristics and the plurality of expression base templates after projection transformation.
In a possible implementation manner, the determining module 433 may be specifically configured to:
performing primary projection transformation on each expression base template by using a preset initial transformation coefficient, and obtaining a plurality of template coefficients contained in the objective function according to a pre-established objective function, the human face features and a plurality of expression base templates subjected to primary projection transformation;
and taking the obtained template coefficients as the updated transformation coefficients of the expression base templates, performing re-projection transformation on the expression base templates by using the updated transformation coefficients, and performing optimization processing on the objective function according to the objective function, the human face features and the plurality of expression base templates after the re-projection transformation until the preset requirements are met.
In a possible implementation manner, the facial feature includes a plurality of facial feature points, each expression base template includes a plurality of expression feature points, and the determining module 433 may be configured to implement projective transformation by:
determining expression feature points corresponding to the facial feature points in each expression base template aiming at each facial feature point in the facial features;
and carrying out projection transformation on each expression base template according to the coordinate values of the expression feature points in each expression base template and the coordinate values of the corresponding face feature points.
In a possible implementation manner, the determining module 433 may be specifically configured to:
obtaining a first array constructed by the plurality of human face feature points, and obtaining a second array constructed by a plurality of expression feature points contained in each expression base template;
and based on a pre-established target function, performing weighted accumulation on the plurality of second arrays according to corresponding template coefficients, subtracting the first array, and determining each template coefficient when the target function meets the preset requirement.
In a possible implementation manner, the control module 434 may specifically be configured to:
carrying out weighted accumulation on the plurality of expression base templates according to corresponding template coefficients to obtain control information;
and performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image according to the control information.
In a possible implementation manner, the control information includes a plurality of pieces of coordinate information, each piece of coordinate information is a weighted result of coordinates of corresponding expressive feature points included in a plurality of expressive base templates, and the control module 434 may be specifically configured to:
acquiring virtual features corresponding to the human face features in the constructed virtual image, and acquiring a plurality of virtual feature points contained in the virtual features;
and controlling each virtual feature point in the virtual features according to the corresponding coordinate information in the control information so as to drive the virtual image to carry out state follow-up.
The description of the processing flow of each module in the device and the interaction flow between the modules may refer to the related description in the above method embodiments, and will not be described in detail here.
In an embodiment of the present application, there is also provided a computer-readable storage medium having stored therein a computer program for executing the steps of the avatar follow-up control method when the computer program runs, corresponding to the avatar follow-up control method described above.
The steps executed when the computer program runs are not described in detail herein, and the explanation of the avatar servo control method may be referred to in the foregoing.
To sum up, according to the virtual image follow-up control method, the virtual image follow-up control device, the electronic device, and the readable storage medium provided by the embodiments of the present application, a human face image of a anchor is obtained, and a plurality of human face features included in the human face image are obtained, wherein for each human face feature, a plurality of preset expression base templates corresponding to the human face feature can be obtained, and then a template coefficient of the human face feature relative to each expression base template is determined according to the human face feature and the plurality of expression base templates. And performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image based on each expression base template and the corresponding template coefficients. According to the scheme, the facial features of the anchor are embodied by the expression base template carrying the template coefficients, so that the virtual image is controlled, follow-up control over the virtual image can be achieved under the condition that other auxiliary equipment is not needed, and the control information is more standard and more accurate in control by controlling the information of the expression base template.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims (11)

1. An avatar follow-up control method, the method comprising:
acquiring a human face image of a anchor, and acquiring a plurality of human face features contained in the human face image;
aiming at each face feature, obtaining a plurality of preset expression base templates corresponding to the face feature;
determining template coefficients of the facial features relative to each expression base template according to the facial features and the expression base templates;
and performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image based on each expression base template and the corresponding template coefficients.
2. The avatar follow-up control method of claim 1, wherein said step of determining template coefficients of said facial features with respect to each of said expression base templates based on said facial features and said plurality of expression base templates comprises:
and optimizing the objective function according to a pre-established objective function, the facial features and the expression base templates until a plurality of template coefficients contained in the objective function are obtained when a preset requirement is met, wherein the template coefficients correspond to the expression base templates one to one.
3. The avatar follow-up control method according to claim 2, wherein said step of performing optimization processing on said objective function according to a pre-established objective function, said facial features and said plurality of expression base templates comprises:
and performing projection transformation on each expression base template, and performing optimization processing on the objective function according to a pre-established objective function, the human face characteristics and the plurality of expression base templates after projection transformation.
4. The avatar follow-up control method of claim 3, wherein said step of performing a projective transformation on each of said expression base templates and performing an optimization process on said objective function according to a pre-established objective function, said facial features and a plurality of expression base templates after the projective transformation comprises:
performing primary projection transformation on each expression base template by using a preset initial transformation coefficient, and obtaining a plurality of template coefficients contained in the objective function according to a pre-established objective function, the human face features and a plurality of expression base templates subjected to primary projection transformation;
and taking the obtained template coefficients as the updated transformation coefficients of the expression base templates, performing re-projection transformation on the expression base templates by using the updated transformation coefficients, and performing optimization processing on the objective function according to the objective function, the human face features and the plurality of expression base templates after the re-projection transformation until the preset requirements are met.
5. The avatar follow-up control method of claim 3, wherein said facial features comprise a plurality of facial feature points, each of said emoji templates comprises a plurality of emoji feature points;
the step of performing projection transformation on each expression base template comprises the following steps:
determining expression feature points corresponding to the facial feature points in each expression base template aiming at each facial feature point in the facial features;
and carrying out projection transformation on each expression base template according to the coordinate values of the expression feature points in each expression base template and the coordinate values of the corresponding face feature points.
6. The avatar follow-up control method of claim 3, wherein said facial features comprise a plurality of facial feature points, each of said emoji templates comprises a plurality of emoji feature points;
the step of optimizing the objective function according to the pre-established objective function, the facial features and the expression base templates after projection transformation comprises the following steps:
obtaining a first array constructed by the plurality of human face feature points, and obtaining a second array constructed by a plurality of expression feature points contained in each expression base template;
and based on a pre-established target function, performing weighted accumulation on the plurality of second arrays according to corresponding template coefficients, subtracting the first array, and determining each template coefficient when the target function meets the preset requirement.
7. The avatar servo-control method of claim 1, wherein said step of performing state servo-control on the virtual features corresponding to said facial features in the constructed avatar based on each said expression base template and its respective corresponding template coefficients comprises:
carrying out weighted accumulation on the plurality of expression base templates according to corresponding template coefficients to obtain control information;
and performing state follow-up control on virtual features corresponding to the human face features in the constructed virtual image according to the control information.
8. The avatar servo-control method of claim 7, wherein said control information includes a plurality of coordinate information, each of said coordinate information being a weighted result of coordinates of corresponding expressive feature points included in a plurality of expressive base templates;
the step of performing state follow-up control on virtual features corresponding to the face features in the constructed virtual image according to the control information comprises the following steps:
acquiring virtual features corresponding to the human face features in the constructed virtual image, and acquiring a plurality of virtual feature points contained in the virtual features;
and controlling each virtual feature point in the virtual features according to the corresponding coordinate information in the control information so as to drive the virtual image to carry out state follow-up.
9. An avatar follow-up control apparatus, the apparatus comprising:
the first acquisition module is used for acquiring a human face image of a anchor and acquiring a plurality of human face features contained in the human face image;
the second obtaining module is used for obtaining a plurality of corresponding expression base templates aiming at each face feature;
the determining module is used for determining a template coefficient of the face feature relative to each expression base template according to the face feature and the expression base templates;
and the control module is used for carrying out state follow-up control on virtual features corresponding to the human face features in the constructed virtual image based on each expression base template and the corresponding template coefficient.
10. An electronic device comprising one or more storage media and one or more processors in communication with the storage media, the one or more storage media storing processor-executable machine-executable instructions that, when executed by the electronic device, are executed by the processors to perform the method steps of any of claims 1-8.
11. A computer-readable storage medium, characterized in that it stores machine-executable instructions which, when executed, implement the method steps of any one of claims 1-8.
CN202110517578.4A 2021-05-12 2021-05-12 Avatar follow-up control method and device, electronic equipment and readable storage medium Pending CN113192164A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110517578.4A CN113192164A (en) 2021-05-12 2021-05-12 Avatar follow-up control method and device, electronic equipment and readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110517578.4A CN113192164A (en) 2021-05-12 2021-05-12 Avatar follow-up control method and device, electronic equipment and readable storage medium

Publications (1)

Publication Number Publication Date
CN113192164A true CN113192164A (en) 2021-07-30

Family

ID=76981250

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110517578.4A Pending CN113192164A (en) 2021-05-12 2021-05-12 Avatar follow-up control method and device, electronic equipment and readable storage medium

Country Status (1)

Country Link
CN (1) CN113192164A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113655889A (en) * 2021-09-01 2021-11-16 北京字跳网络技术有限公司 Virtual role control method and device and computer storage medium
CN114245155A (en) * 2021-11-30 2022-03-25 北京百度网讯科技有限公司 Live broadcast method and device and electronic equipment
CN116112716A (en) * 2023-04-14 2023-05-12 世优（北京）科技有限公司 Virtual person live broadcast method, device and system based on single instruction stream and multiple data streams

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113655889A (en) * 2021-09-01 2021-11-16 北京字跳网络技术有限公司 Virtual role control method and device and computer storage medium
CN113655889B (en) * 2021-09-01 2023-08-08 北京字跳网络技术有限公司 Virtual character control method, device and computer storage medium
CN114245155A (en) * 2021-11-30 2022-03-25 北京百度网讯科技有限公司 Live broadcast method and device and electronic equipment
CN116112716A (en) * 2023-04-14 2023-05-12 世优（北京）科技有限公司 Virtual person live broadcast method, device and system based on single instruction stream and multiple data streams
CN116112716B (en) * 2023-04-14 2023-06-09 世优（北京）科技有限公司 Virtual person live broadcast method, device and system based on single instruction stream and multiple data streams

Similar Documents

Publication Publication Date Title
CN109325933B (en) Method and device for recognizing copied image
CN113192164A (en) Avatar follow-up control method and device, electronic equipment and readable storage medium
CN109034013B (en) Face image recognition method, device and storage medium
CN111242097B (en) Face recognition method and device, computer readable medium and electronic equipment
EP3709266A1 (en) Human-tracking methods, apparatuses, systems, and storage media
US8903139B2 (en) Method of reconstructing three-dimensional facial shape
WO2022041830A1 (en) Pedestrian re-identification method and device
WO2018133825A1 (en) Method for processing video images in video call, terminal device, server, and storage medium
CN111160202B (en) Identity verification method, device, equipment and storage medium based on AR equipment
JP2015096812A (en) Image processor, imaging device and distance correction method
CN111008935B (en) Face image enhancement method, device, system and storage medium
US20220214797A1 (en) Virtual image control method, apparatus, electronic device and storage medium
JP7342366B2 (en) Avatar generation system, avatar generation method, and program
CN113887547B (en) Key point detection method and device and electronic equipment
CN111914811B (en) Image data processing method, image data processing device, computer equipment and storage medium
CN108388889B (en) Method and device for analyzing face image
CN113313097B (en) Face recognition method, terminal and computer readable storage medium
CN112543343A (en) Live broadcast picture processing method and device based on live broadcast with wheat and electronic equipment
CN113422977A (en) Live broadcast method and device, computer equipment and storage medium
CN112562056A (en) Control method, device, medium and equipment for virtual light in virtual studio
US20230206093A1 (en) Music recommendation method and apparatus
CN110149476A (en) A kind of time-lapse photography method, apparatus, system and terminal device
CN109089058B (en) Video picture processing method, electronic terminal and device
CN112446254A (en) Face tracking method and related device
CN113902932A (en) Feature extraction method, visual positioning method and device, medium and electronic equipment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination