WO2020056903A1 - Procédé et dispositif de génération d'informations - Google Patents

Procédé et dispositif de génération d'informations Download PDF

Info

Publication number
WO2020056903A1
WO2020056903A1 PCT/CN2018/115974 CN2018115974W WO2020056903A1 WO 2020056903 A1 WO2020056903 A1 WO 2020056903A1 CN 2018115974 W CN2018115974 W CN 2018115974W WO 2020056903 A1 WO2020056903 A1 WO 2020056903A1
Authority
WO
WIPO (PCT)
Prior art keywords
face detection
detection frame
position information
weight
target
Prior art date
Application number
PCT/CN2018/115974
Other languages
English (en)
Chinese (zh)
Inventor
吴兴龙
Original Assignee
北京字节跳动网络技术有限公司
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 北京字节跳动网络技术有限公司 filed Critical 北京字节跳动网络技术有限公司
Publication of WO2020056903A1 publication Critical patent/WO2020056903A1/fr

Links

Images

Classifications

    • 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
    • G06V40/161—Detection; Localisation; Normalisation
    • 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
    • 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
    • G06V40/172—Classification, e.g. identification

Definitions

  • Embodiments of the present application relate to the field of computer technology, and in particular, to a method and an apparatus for generating information.
  • Face detection refers to the process of searching for a given image using a certain strategy to determine whether it contains a face object, and if so, returning the position and size of the face object, and The returned result can be reflected in the form of a face detection frame in the image.
  • a related method is to directly perform face detection on each frame to obtain a face detection frame for indicating a face object in each frame.
  • the embodiments of the present application provide a method and device for generating information.
  • an embodiment of the present application provides a method for generating information.
  • the method includes: obtaining position information of a first face detection frame obtained by performing face detection on a current frame of a target video in advance, and To obtain pre-stored position information of the second face detection frame in a previous frame of the current frame; based on the obtained position information, determine the intersection ratio of the first face detection frame and the second face detection frame; based on The intersection ratio determines the weight of the obtained position information of each face detection frame; based on the determined weight and the obtained position information, determines the target position information of the first face detection frame to detect the first face The position of the box is updated.
  • determining the weight of the obtained position information of each face detection frame based on the intersection ratio includes: performing the power operation with the intersection ratio as the base and the first preset value as the exponent; The calculation result of the operation is determined as the weight of the position information of the second face detection frame, and the difference between the second preset value and the weight is determined as the weight of the position information of the first face detection frame.
  • determining the weight of the obtained position information of each face detection frame based on the intersection ratio includes: using a natural constant as a base number, and taking a difference between a reciprocal of the intersection ratio and a second preset value as an index , Performing a power operation; determining the inverse of the calculation result of the power operation as the weight of the position information of the second face detection frame, and determining the difference between the second preset value and the weight as the weight of the position information of the first face detection frame .
  • the position information of the first face detection frame includes designated diagonal vertex coordinates of the first face detection frame
  • the position information of the second face detection frame includes designated diagonal vertices of the second face detection frame. Coordinates; and determining target position information of the first face detection frame based on the determined weight and the obtained position information to update the position of the first face detection frame, including: updating the position of the first face detection frame
  • the weight of the position information is used as the weight of the specified diagonal vertex coordinates of the first face detection frame
  • the weight of the position information of the second face detection frame is used as the weight of the specified diagonal vertex coordinates of the second face detection frame.
  • the weighted calculation result of the designated diagonal vertex coordinates of a face detection frame and the designated diagonal vertex coordinates of a second face detection frame is determined as the target diagonal vertex coordinates of the first face detection frame to detect the first face.
  • the position of the box is updated.
  • the designated diagonal vertex coordinates of the first face detection frame include a first vertex coordinate and a second vertex coordinate
  • the designated diagonal vertex coordinates of the second face detection frame include a third vertex coordinate and a fourth vertex. Coordinates; and determining the weighted calculation result of the designated diagonal vertex coordinates of the first face detection frame and the designated diagonal vertex coordinates of the second face detection frame as the target diagonal vertex coordinates of the first face detection frame, including: The weighted calculation result of the abscissa of the first vertex coordinate and the abscissa of the third vertex coordinate is determined as the first target abscissa; the weighted calculation result of the ordinate of the first vertex coordinate and the ordinate of the third vertex coordinate is determined as The first target ordinate; the weighted calculation result of the abscissa of the second vertex coordinate and the abscissa of the fourth vertex coordinate is determined as the second target abscissa; the ordinate of the second vertex
  • an embodiment of the present application provides an apparatus for generating information.
  • the apparatus includes: an obtaining unit configured to obtain first face detection obtained by performing face detection on a current frame of a target video in advance; Position information of the frame, and acquiring position information of a second face detection frame in a previous frame of a current frame that is stored in advance; a first determining unit configured to determine the first face detection based on the acquired position information The intersection ratio of the frame and the second face detection frame; the second determination unit is configured to determine the weight of the position information of each face detection frame obtained based on the intersection ratio; the update unit is configured to be based on the determined Determine the target position information of the first face detection frame to update the position of the first face detection frame.
  • the second determination unit includes: a first operation module configured to perform a power operation with a cross-ratio as a base and a first preset value as an exponent; the first determination module is configured to convert The calculation result of the power operation is determined as the weight of the position information of the second face detection frame, and the difference between the second preset value and the weight is determined as the weight of the position information of the first face detection frame.
  • the second determining unit includes: a second operation module configured to perform a power operation with a natural constant as a base, and a difference between a reciprocal of the intersection ratio and a second preset value as an exponent; a second The determining module is configured to determine the inverse of the calculation result of the power operation as the weight of the position information of the second face detection frame, and determine the difference between the second preset value and the weight as the position information of the first face detection frame. Weights.
  • the position information of the first face detection frame includes designated diagonal vertex coordinates of the first face detection frame
  • the position information of the second face detection frame includes designated diagonal vertices of the second face detection frame.
  • Coordinates; and an update unit further configured to: use the weight of the position information of the first face detection frame as the weight of the designated diagonal vertex coordinates of the first face detection frame, and set the position information of the second face detection frame to The weight is the weight of the designated diagonal vertex coordinates of the second face detection frame, and the weighted calculation result of the designated diagonal vertex coordinates of the first face detection frame and the designated diagonal vertex coordinates of the second face detection frame is determined as The coordinates of the diagonal diagonal vertices of a face detection frame to update the position of the first face detection frame.
  • the designated diagonal vertex coordinates of the first face detection frame include a first vertex coordinate and a second vertex coordinate
  • the designated diagonal vertex coordinates of the second face detection frame include a third vertex coordinate and a fourth vertex.
  • an update unit further configured to: determine a weighted calculation result of the abscissa of the first vertex coordinate and the abscissa of the third vertex coordinate as the first target abscissa; and determine the ordinate of the first vertex coordinate and the third The weighted calculation result of the ordinate of the vertex coordinate is determined as the first target ordinate; the weighted calculation result of the abscissa of the second vertex coordinate and the abscissa of the fourth vertex coordinate is determined as the second target abscissa; the second vertex coordinate The weighted calculation result of the ordinate of the ordinate of the fourth coordinate and the ordinate of the fourth vertex is determined as the second target ordinate; the coordinates formed by the first target abscis
  • an embodiment of the present application provides an electronic device including: one or more processors; a storage device that stores one or more programs thereon; when one or more programs are processed by one or more processors Execution causes one or more processors to implement the method as in any one of the first aspects described above.
  • an embodiment of the present application provides a computer-readable medium on which a computer program is stored, and when the program is executed by a processor, the method as in any one of the foregoing first embodiments is implemented.
  • the method and device for generating information provided in the embodiments of the present application are performed by performing position information on a first face detection frame of a current frame of a target video and position information on a second face detection frame of a previous frame. Obtained, so that the intersection ratio of the first face detection frame and the second face detection frame can be determined based on the obtained position information. After that, based on the intersection ratio, the weight of the obtained position information of each face detection frame is determined, and the weight of the obtained position information of each face detection frame can be determined. Finally, the target position information of the first face detection frame may be determined based on the determined weight and the obtained position information to update the position of the first face detection frame.
  • the position of the face detection frame in the subsequent frames can be adjusted based on the intersection of the face detection frames in the two frames before and after.
  • the position of the face detection frame in the subsequent frame considers the position of the face detection frame in the previous frame, and the entire area of the face detection frame in the previous frame is considered instead of a single coordinate, thereby reducing the
  • the jitter of the face detection frame improves the smoothness and stability of the movement of the face detection frame in the video.
  • FIG. 1 is an exemplary system architecture diagram to which an embodiment of the present application can be applied;
  • FIG. 1 is an exemplary system architecture diagram to which an embodiment of the present application can be applied;
  • FIG. 2 is a flowchart of an embodiment of a method for generating information according to the present application
  • FIG. 3 is a schematic diagram of an application scenario of a method for generating information according to the present application.
  • FIG. 4 is a flowchart of still another embodiment of a method for generating information according to the present application.
  • FIG. 5 is a schematic structural diagram of an embodiment of an apparatus for generating information according to the present application.
  • FIG. 6 is a schematic structural diagram of a computer system suitable for implementing an electronic device according to an embodiment of the present application.
  • FIG. 1 illustrates an exemplary system architecture 100 to which a method for detecting a key point of a face or an apparatus for detecting a key point of a face of the present application can be applied.
  • the system architecture 100 may include terminal devices 101, 102, and 103, a network 104, and a server 105.
  • the network 104 is a medium for providing a communication link between the terminal devices 101, 102, 103 and the server 105.
  • the network 104 may include various types of connections, such as wired, wireless communication links, or fiber optic cables, and so on.
  • the user can use the terminal devices 101, 102, 103 to interact with the server 105 through the network 104 to receive or send messages and the like.
  • Various communication client applications can be installed on the terminal devices 101, 102, and 103, such as voice interaction applications, shopping applications, search applications, instant communication tools, email clients, social platform software, and the like.
  • the terminal devices 101, 102, and 103 may be hardware or software.
  • the terminal devices 101, 102, and 103 can be various electronic devices that have a display screen and support web browsing, including but not limited to smartphones, tablets, e-book readers, MP3 players (Moving Pictures Experts Group) Audio Layer III, moving picture expert compression standard audio layer 3), MP4 (Moving Picture Experts Group Audio, Layer 4 IV, moving picture expert compression standard audio layer 4) player, laptop portable computer and desktop computer, etc.
  • MP3 players Motion Pictures Experts Group Audio Layer III, moving picture expert compression standard audio layer 3
  • MP4 Motion Picture Experts Group Audio, Layer 4 IV, moving picture expert compression standard audio layer 4
  • player laptop portable computer and desktop computer, etc.
  • laptop portable computer and desktop computer etc.
  • the terminal devices 101, 102, and 103 are software, they can be installed in the electronic devices listed above. It can be implemented as multiple software or software modules (for example, to provide distributed services), or it can be implemented as a single software or software module. It is not specifically limited
  • an image acquisition device may also be installed thereon.
  • the image acquisition device can be various devices that can implement the function of acquiring images, such as cameras, sensors, and so on. Users can use the image capture device on the terminal devices 101, 102, 103 to capture video.
  • the terminal device 101, 102, 103 can perform face detection and other processing on the video that it plays or frames recorded by the user; it can also analyze the face detection results (such as the position information of the face detection frame), etc. Process and update the position of the face detection frame.
  • the server 105 may be a server providing various services, such as a video processing server for storing, managing, or analyzing videos uploaded by the terminal devices 101, 102, and 103.
  • the video processing server can store a large number of videos, and can send videos to the terminal devices 101, 102, and 103.
  • the server 105 may be hardware or software.
  • the server can be implemented as a distributed server cluster consisting of multiple servers or as a single server.
  • the server can be implemented as multiple software or software modules (for example, to provide distributed services), or it can be implemented as a single software or software module. It is not specifically limited here.
  • the methods for generating information provided by the embodiments of the present application are generally executed by the terminal devices 101, 102, and 103. Accordingly, the devices for generating information are generally provided in the terminal devices 101, 102, and 103.
  • the server 105 may not be provided in the system architecture 100.
  • the server 105 may also perform face detection and other processing on its stored videos or videos uploaded by the terminal devices 101, 102, and 103, and return the processing results to the terminal devices 101, 102, and 103.
  • the method for generating information provided in the embodiment of the present application may also be executed by the server 105, and accordingly, the apparatus for generating information may also be set in the server 105.
  • terminal devices, networks, and servers in FIG. 1 are merely exemplary. According to implementation needs, there can be any number of terminal devices, networks, and servers.
  • a flowchart 200 of one embodiment of a method for generating information according to the present application is shown.
  • the method for generating information includes the following steps:
  • Step 201 Obtain position information of a first face detection frame obtained by performing face detection on a current frame of a target video in advance, and obtain a second face detection frame in a previous frame of a current frame that is stored in advance. location information.
  • an execution subject of the method for generating information may record or play a video.
  • the video that it plays may be a video that is stored locally in advance; it may also be a video that is obtained from a server (such as the server 105 shown in FIG. 1) through a wired connection or a wireless connection.
  • a server such as the server 105 shown in FIG. 1
  • the above-mentioned execution body may be installed or connected with an image acquisition device (for example, a camera).
  • wireless connection methods may include, but are not limited to, 3G / 4G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods now known or developed in the future. .
  • the execution subject may obtain position information of a first face detection frame obtained by performing face detection on a current frame of a target video in advance, and obtain a previously stored previous frame of the current frame. Position information of the second face detection frame.
  • the target video may be a video currently being played or a video being recorded by a user. It is not limited here.
  • the current frame of the target video may be a frame in the target video whose position detection frame is to be updated.
  • the above-mentioned execution subject may perform face detection on each frame of the target video in sequence according to the timestamp order of the frames. After performing face detection on each frame except the first frame, The obtained face detection frame is subjected to position correction.
  • the current frame to be corrected for the position of the face detection frame can be referred to as the current frame of the target video. Take the following two scenarios as examples:
  • the target video may be a video being played by the execution subject.
  • the execution subject may perform face detection on each frame to be played one by one to obtain the position information of the face detection frame of the frame.
  • the position information of the face detection frame of the frame may be corrected to further play the frame.
  • the frame at which the position of the face detection frame is about to be corrected at the current moment may be the current frame.
  • the target video may be a video being recorded by the above-mentioned execution subject.
  • the execution subject may perform face detection on each captured frame one by one to obtain the position information of the face detection frame of the frame. After the first frame is captured, for each frame captured next, after performing face detection on the frame, the obtained face detection frame can be position corrected, and then the frame is displayed.
  • the latest frame acquired at the current time and which has not been subjected to the position correction of the face detection frame may be the current frame.
  • a pre-trained face detection model may be stored in the execution subject.
  • the execution subject may input a frame in a frame in the target video into a pre-trained face detection model, and obtain position information of the face detection frame of the frame.
  • the above-mentioned face detection model may be used to detect an area where a face object is located in the image (it may be represented by a face detection frame, and here, the face detection frame may be a rectangular frame).
  • the face detection detection model can output the position information of the face detection frame.
  • the face detection model may be obtained by performing supervised training on an existing convolutional neural network based on a sample set (including a face image and a label for indicating the position of a face object region) using a machine learning method.
  • the convolutional neural network can use various existing structures, such as DenseBox, VGGNet, ResNet, SegNet, and so on. It should be noted that the above-mentioned machine learning method and supervised training method are well-known technologies that are widely studied and applied at present, and will not be repeated here.
  • the position information of the face detection frame may be information for indicating and uniquely determining the position of the face detection frame in the frame.
  • the position information of the face detection frame may include coordinates of four vertices of the face detection frame.
  • the position information of the face detection frame may include the coordinates of any set of diagonal vertices of the face detection frame. For example, the coordinates of the upper left vertex and the coordinates of the lower right vertex.
  • the position information of the face detection frame may include the coordinates of any vertex of the face detection frame and the length and width of the face detection frame.
  • position information is not limited to the above list, and may also include other information that can be used to indicate and uniquely determine the position of the face detection frame.
  • Step 202 Determine an intersection ratio of the first face detection frame and the second face detection frame based on the obtained position information.
  • the execution subject may determine the first face detection frame and the second face based on the obtained position information of the first face detection frame and the position information of the second face detection frame.
  • the intersection ratio of two rectangles may be the ratio of the area of the area where the two rectangles intersect to the area of the area where the two rectangles intersect.
  • the area of the area where the two rectangles meet is equal to the sum of the areas of the two rectangles minus the area of the area where the two rectangles intersect.
  • the intersection ratio is a number in the interval [0,1].
  • the position of the face detection frame in the frame can be determined. Therefore, by using the position information of the first face detection frame, the coordinates of each vertex of the first face detection frame in the current frame can be determined. Based on the position information of the second face detection frame, the coordinates of each vertex of the second face detection frame in the previous frame of the current frame can be determined.
  • the position information of the face detection frame can include the coordinates of any vertex (such as the upper left vertex) of the face detection frame and the length and width of the face detection frame
  • the horizontal coordinate of the upper left vertex can be related to the length Add and add the vertical coordinates of the upper left vertex and the width to obtain the coordinates of the upper right vertex, the lower left vertex, and the lower right vertex, respectively.
  • the coordinates of each vertex of the first face detection frame and the coordinates of each vertex of the second face detection frame can be obtained. Therefore, the vertex coordinates of the first face detection frame and the vertex coordinates of the second face detection frame can be used to determine the length and width of the rectangle where the first face detection frame and the second face detection frame intersect. Further, the area of the intersecting rectangles (may be referred to as an intersecting area) can be obtained. After that, the sum of the areas of the first face detection frame and the second face detection frame (which can be referred to as the total area) can be calculated. Then, the difference between the total area and the intersecting area can be calculated (may be referred to as the merging area). Finally, the ratio of the intersection area and the merging area can be determined as the intersection ratio of the first face detection frame and the second face detection frame.
  • Step 203 Determine the weight of the obtained position information of each face detection frame based on the intersection and union ratio.
  • the execution subject may determine the weight of the position information of the first face detection frame and the weight of the second face detection frame based on the intersection ratio determined in step 202, respectively.
  • the execution subject may determine the weight of the position information of the first face detection frame and the weight of the second face detection frame based on the intersection ratio determined in step 202, respectively.
  • the intersection ratio can be calculated in a formula established in advance, and the calculation result is determined as the weight of the position information of the second face detection frame.
  • the foregoing formulas established in advance may be various formulas that satisfy preset conditions, and are not limited herein.
  • the above preset conditions include: the larger the cross-ratio, the larger the calculation result of the above formula; the smaller the cross-ratio, the smaller the calculation result of the above formula.
  • the difference between the preset value (for example, 1) and the weight of the position information of the second face detection frame may be determined as the weight of the position information of the first face detection frame.
  • the order of determining the weight of the position information of the first face detection frame and the weight of the position information of the second face detection frame is not limited herein.
  • the execution subject may modify the pre-established formula so as to first determine the weight of the position information of the first face detection frame, and then determine the weight of the position information of the second face detection frame.
  • the above-mentioned execution body may perform the power operation with the above-mentioned intersection ratio as a base and a first preset value (for example, 6, or 3, etc.) as an index.
  • the first preset value may be determined by a technician based on a large amount of data statistics and experiments.
  • the execution subject may determine the calculation result of the power operation as the weight of the position information of the second face detection frame, and combine the second preset value (for example, 1) with the determined position information of the second face detection frame. The difference between the weights is determined as the weight of the position information of the first face detection frame.
  • the execution body may perform a power operation by using a natural constant as a base and a difference between the reciprocal of the intersection ratio and a second preset value (for example, 1) as an index. Then, the inverse of the calculation result of the power operation may be determined as the weight of the position information of the second face detection frame, and the difference between the second preset value and the weight of the determined position information of the second face detection frame. Determined as the weight of the position information of the first face detection frame.
  • a second preset value for example, 1
  • the above-mentioned execution body may also determine the weight of each position information obtained in other ways. It is not limited to the above implementation. For example, a certain preset value (for example, 2 or 3) can be used as the base, and the difference between the reciprocal of the above-mentioned intersection ratio and the second preset value (for example, 1) can be used as an index to perform a power operation. Then, the inverse of the calculation result of the power calculation may be determined as the weight of the position information of the second face detection frame, and the difference between the second preset value and the weight may be determined as the position information of the first face detection frame. the weight of.
  • a certain preset value for example, 2 or 3
  • the difference between the reciprocal of the above-mentioned intersection ratio and the second preset value for example, 1
  • the inverse of the calculation result of the power calculation may be determined as the weight of the position information of the second face detection frame
  • the difference between the second preset value and the weight may be determined as the position information of the first face detection
  • the average value of the coordinates of the corresponding vertex (for example, the upper-left vertex) in the face detection frame in the previous frame and the current frame is usually used as the corrected coordinate of the vertex (upper-left vertex) in the current frame .
  • the corrected coordinates of each vertex of the current frame are thus obtained.
  • the position of the face detection frame in the current frame is corrected by using the weight determined based on the crossover ratio in this application. Because the larger the intersection ratio, the slower the face object movement; the smaller the intersection ratio, the faster the face object movement. Therefore, different weights can be calculated according to different cross-ratio ratios. Thereby, the drag feeling is reduced, and the timeliness and accuracy of the face detection frame are improved.
  • the conventional method there is also a method for determining the weight of the coordinates of the vertex of the face detection frame in the previous frame and the distance between the coordinates of corresponding vertices in the face detection frame (for example, all of them are the upper left vertex).
  • the weights of the coordinates of each vertex are independent, and the entire face detection frame cannot be considered. Therefore, the smoothing effect is poor.
  • the entire area of the entire face detection frame is considered in the process of determining the intersection ratio, and the weights of the coordinates of each vertex in the same face detection frame are the same, so Consider the face detection frame as a whole. Improved smoothing effect.
  • Step 204 Determine target position information of the first face detection frame based on the determined weight and the obtained position information to update the position of the first face detection frame.
  • the above-mentioned execution subject may determine the target position information of the first face detection frame based on the determined weight and the obtained position information to update the position of the first face detection frame.
  • the execution subject may modify the position information of the first face detection frame based on the determined weight. That is, the vertex coordinates of the first face detection frame are corrected.
  • the position information of the face detection frame may include coordinates of four vertices of the face detection frame.
  • the execution subject may modify the coordinates of the first face detection frame, respectively. Specifically, for each vertex, the following steps can be performed (here, the sitting vertex is used as an example for description, and the remaining vertices will not be described again):
  • the abscissa of the upper left vertex of the first face detection frame and the abscissa of the upper left vertex of the second face detection frame are weighted. That is, the horizontal coordinate of the upper left vertex of the first face detection frame is multiplied by the weight of the position information of the first face detection frame to obtain a first value. Multiply the abscissa of the upper left vertex of the second face detection frame by the weight of the position information of the second face detection frame to obtain a second value. The product of the first value and the second value is determined as the abscissa of the upper left vertex of the first face detection frame after correction.
  • the vertical coordinate of the upper left vertex of the first face detection frame and the vertical coordinate of the upper left vertex of the second face detection frame are weighted. That is, the vertical coordinate of the upper left vertex of the first face detection frame is multiplied by the weight of the position information of the first face detection frame to obtain a third value. Multiply the vertical coordinate of the upper left vertex of the second face detection frame by the weight of the position information of the second face detection frame to obtain a fourth value. The product of the third value and the fourth value is determined as the vertical coordinate of the upper left vertex of the first face detection frame after correction.
  • the abscissa and ordinate obtained in the first step and the second step are respectively aggregated into the coordinates of the upper left vertex of the first face detection frame after correction.
  • the above-mentioned electronic device may summarize the coordinates of the corrected vertices as target position information. Therefore, the position of the first face detection frame can be updated.
  • the position information of the first face detection frame includes the specified diagonal vertex coordinates of the first face detection frame
  • the position information of the second face detection frame includes the first The specified diagonal vertex coordinates of the two-face detection frame.
  • the designated diagonal vertex coordinates of the first face detection frame may include coordinates of a first vertex (for example, an upper left vertex) and coordinates of a second vertex (for example, a lower right vertex).
  • the designated diagonal vertex coordinates of the second face detection frame may include coordinates of a third vertex (for example, an upper left vertex) and coordinates of a fourth vertex (for example, a lower right vertex).
  • the execution subject may use the weight of the position information of the first face detection frame as the weight of the specified diagonal vertex coordinates of the first face detection frame, and the weight of the position information of the second face detection frame.
  • a weight of the designated diagonal vertex coordinates of the second face detection frame a weighted calculation result of the designated diagonal vertex coordinates of the first face detection frame and the designated diagonal vertex coordinates of the second face detection frame is determined.
  • the coordinates of the diagonal vertices of the target of the first face detection frame to update the position of the first face detection frame.
  • the target diagonal vertex coordinates of the first face detection frame can be calculated in the following sequence of operations:
  • a weighted calculation result of the abscissa of the first vertex coordinate and the abscissa of the third vertex coordinate may be determined as the first target abscissa;
  • a weighted calculation result of the ordinate of the first vertex coordinate and the ordinate of the third vertex coordinate may be determined as the first target ordinate;
  • a weighted calculation result of the abscissa of the second vertex coordinate and the abscissa of the fourth vertex coordinate may be determined as the second target abscissa;
  • the weighted calculation result of the ordinate of the second vertex coordinate and the ordinate of the fourth vertex coordinate may be determined as the second target ordinate;
  • the coordinates formed by the first target abscissa and the first target ordinate, and the coordinates formed by the second target abscissa and the second target ordinate can be determined as the targets of the first face detection frame.
  • Diagonal vertex coordinates Since the coordinates of a set of diagonal apex points are known, the position of the rectangular frame can be uniquely determined. Thereby, the position of the first face detection frame can be updated.
  • another set of diagonal vertex coordinates of the first face detection frame may be calculated according to the target diagonal vertex coordinates. Thereby, the coordinates of the four vertices of the first face detection frame are obtained.
  • the position information of the face detection frame may include the coordinates of any vertex of the face detection frame and the length and width of the face detection frame.
  • the execution subject may first determine the coordinates of a diagonal vertex of the vertex based on the coordinates of the vertex, the length, and the width. Alternatively, determine the coordinates of the remaining three vertices. Then, the target position information of the first face detection frame can be determined by using the operation steps described in the above two implementation manners. Thereby, the position of the first face detection frame is updated.
  • FIG. 3 is a schematic diagram of an application scenario of the method for generating information according to this embodiment.
  • a user uses a self-timer mode of the terminal device 301 to record a target video.
  • the terminal device After capturing the first frame, uses the stored face detection model to perform face point detection on the first frame, and obtains the position information 302 of the face detection frame in the first frame.
  • the terminal device After the terminal device captures the second frame, it uses the stored face detection model to perform face detection on the second frame. Then, the position information 303 of the face detection frame of the second frame is obtained. At the same time, the position information 302 of the face detection frame in the first frame is obtained. Then, based on the position information 302 and the position information 303, an intersection ratio of the first face detection frame and the second face detection frame may be determined. After that, the weight of the obtained position information of each face detection frame can be determined based on the intersection ratio, and the weight of the obtained position information 302 and the weight of the position information 303 can be determined. Finally, the target position information 304 of the face detection frame of the second frame (that is, the final position information of the face detection frame of the second frame) may be determined based on the determined weight and the obtained position information 302 and position information 303. .
  • the terminal device After capturing the third frame, the terminal device uses the stored face detection model to perform face point detection on the third frame. Then, the position information 305 of the face detection frame of the third frame is acquired. At the same time, the position information (that is, the target position information 304) of the face detection frame in the updated second frame is acquired. Then, based on the target position information 304 and the position information 305, the intersection ratio of the second face detection frame and the third face detection frame may be determined. After that, the weights of the obtained position information of each face detection frame can be determined based on the intersection ratio, and the weights of the obtained target position information 304 and the weights of the position information 305 can be determined. Finally, the target position information 306 of the face detection frame of the third frame (that is, the final position information of the face detection frame of the third frame may be determined based on the determined weight and the obtained target position information 304 and position information 305). ).
  • the terminal device 301 can obtain the position information of the face detection frame in each frame in the recorded video.
  • the method provided by the foregoing embodiment of the present application obtains the position information of the first face detection frame of the current frame of the target video and the position information of the second face detection frame of the previous frame by the pre-generated target video, so that it can be based on
  • the obtained position information determines an intersection ratio of the first face detection frame and the second face detection frame.
  • the weight of the obtained position information of each face detection frame is determined, and the weight of the obtained position information of each face detection frame can be determined.
  • the target position information of the first face detection frame may be determined based on the determined weight and the obtained position information to update the position of the first face detection frame.
  • the position of the face detection frame in the subsequent frames can be adjusted based on the intersection of the face detection frames in the two frames before and after. Because the position of the face detection frame in the subsequent frame considers the position of the face detection frame in the previous frame, and the entire area of the face detection frame in the previous frame is considered instead of a single coordinate, thereby reducing the The jitter of the face detection frame improves the smoothing effect and movement stability of the face detection frame in the video.
  • a flowchart 400 of yet another embodiment of a method for generating information is shown.
  • the process 400 of the method for generating information includes the following steps:
  • Step 401 Obtain position information of a first face detection frame obtained by performing face detection on a current frame of a target video in advance, and obtain a second face detection frame in a previous frame of a current frame that is stored in advance. location information.
  • an execution subject of the method for generating information may obtain a first face obtained by performing face detection on a current frame of a target video in advance.
  • the position information of the detection frame, and the position information of the second face detection frame obtained by performing face detection on the previous frame of the current frame in advance.
  • the position information of the first face detection frame may include designated diagonal vertex coordinates (such as the coordinates of the upper-left vertex and the lower-right vertex) of the first face detection frame, and the second face detection frame.
  • the position information of can include the designated diagonal vertex coordinates of the second face detection frame.
  • Step 402 Determine the intersection ratio of the first face detection frame and the second face detection frame based on the obtained position information.
  • the above-mentioned executing subject can determine the coordinates of the remaining vertices of the first face detection frame in the current frame by using the position information of the first face detection frame, so that the first face detection frame can be obtained.
  • the coordinates of each vertex can be obtained.
  • the vertex coordinates of the first face detection frame and the vertex coordinates of the second face detection frame can be used to determine the length and width of the rectangle where the first face detection frame and the second face detection frame intersect.
  • the area of the intersecting rectangles (may be referred to as an intersecting area) can be obtained.
  • the sum of the areas of the first face detection frame and the second face detection frame (which can be referred to as the total area) can be calculated.
  • the difference between the total area and the intersecting area can be calculated (may be referred to as the merging area).
  • the ratio of the intersection area and the merging area can be determined as the intersection ratio of the first face detection frame and the second face detection frame.
  • Step 403 Perform a power operation using the natural constant as a base and the difference between the reciprocal of the intersection and the second preset value as an index.
  • the execution body may perform a power operation by using a natural constant as a base and a difference between the reciprocal of the intersection ratio and a second preset value (for example, 1) as an index.
  • Step 404 Determine the inverse of the calculation result of the power operation as the weight of the position information of the second face detection frame, and determine the difference between the second preset value and the weight as the weight of the position information of the first face detection frame.
  • the execution subject may determine the inverse of the calculation result of the power operation as the weight of the position information of the second face detection frame, and combine the second preset value (for example, 1) with the determined weight. The difference is determined as the weight of the position information of the first face detection frame.
  • Step 405 Use the weight of the position information of the first face detection frame as the weight of the designated diagonal vertex coordinates of the first face detection frame, and use the weight of the position information of the second face detection frame as the second face detection frame.
  • the weight of the specified diagonal vertex coordinates of the first face detection frame, and the weighted calculation result of the specified diagonal vertex coordinates of the first face detection frame and the specified diagonal vertex coordinates of the second face detection frame is determined as the target pair of the first face detection frame Angular vertex coordinates to update the position of the first face detection frame.
  • the execution subject may use the weight of the position information of the first face detection frame as the weight of the designated diagonal vertex coordinates of the first face detection frame.
  • the weighted calculation result of the designated diagonal vertex coordinates of the first face detection frame and the designated diagonal vertex coordinates of the second face detection frame is determined as the target diagonal vertex coordinates of the first face detection frame to
  • the position of the face detection frame is updated.
  • the designated diagonal vertex coordinates of the first face detection frame may include coordinates of a first vertex (for example, an upper left vertex) and coordinates of a second vertex (for example, a lower right vertex).
  • the designated diagonal vertex coordinates of the second face detection frame may include coordinates of a third vertex (for example, an upper left vertex) and coordinates of a fourth vertex (for example, a lower right vertex).
  • the weighted calculation result of the abscissa of the first vertex coordinate and the abscissa of the third vertex coordinate may be determined as the first target abscissa first. Then, a weighted calculation result of the ordinate of the first vertex coordinate and the ordinate of the third vertex coordinate may be determined as the first target ordinate. Then, a weighted calculation result of the abscissa of the second vertex coordinate and the abscissa of the fourth vertex coordinate may be determined as the second target abscissa. Then, a weighted calculation result of the ordinate of the second vertex coordinate and the ordinate of the fourth vertex coordinate may be determined as the second target ordinate.
  • the coordinates formed by the first target abscissa and the first target ordinate, and the coordinates formed by the second target abscissa and the second target ordinate can be determined as the targets of the first face detection frame.
  • Diagonal vertex coordinates Since the coordinates of a set of diagonal vertices are known, the position of the rectangular frame can be uniquely determined. Thereby, the position of the first face detection frame can be updated.
  • the process 400 of the method for generating information in this embodiment highlights the weighting steps of determining the face detection frame of the current frame and the previous frame, respectively.
  • the weight of the position information of the first face detection frame (the face detection frame of the current frame) is larger, and the second face detection frame (the person of the previous frame) Face detection frame) has a smaller weight.
  • the weight of the position information of the first face detection frame is small, and the weight of the position information of the second face detection frame is large. Thereby, the face detection frame can be moved smoothly, the jitter of the face detection frame in the video is further reduced, and the smoothness effect and movement stability of the face detection frame in the video are further improved.
  • this application provides an embodiment of an apparatus for generating information.
  • the apparatus embodiment corresponds to the method embodiment shown in FIG. 2.
  • the device can be specifically applied to various electronic devices.
  • the apparatus 500 for generating information includes: an obtaining unit 501 configured to obtain a first face detection frame obtained by performing face detection on a current frame of a target video in advance; The position information of the second face detection frame in the previous frame of the current frame, and the first determination unit 502 is configured to determine the first person based on the obtained position information.
  • the foregoing second determination unit 503 may include a first operation module and a first determination module (not shown in the figure).
  • the first operation module may be configured to perform the power operation using the intersection ratio as a base and a first preset value as an exponent.
  • the first determination module may be configured to determine a calculation result of the power operation as a weight of the position information of the second face detection frame, and determine a difference between the second preset value and the weight as the first face detection. The weight of the box's position information.
  • the foregoing second determination unit 503 may include a second operation module and a second determination module (not shown in the figure).
  • the second operation module may be configured to perform a power operation using a natural constant as a base and a difference between a reciprocal of the intersection ratio and a second preset value as an index.
  • the second determining module may be configured to determine a reciprocal of a power operation calculation result as a weight of the position information of the second face detection frame, and determine a difference between the second preset value and the weight as the first person. Weight of the position information of the face detection frame.
  • the position information of the first face detection frame may include designated diagonal vertex coordinates of the first face detection frame
  • the position information of the second face detection frame may be Including the designated diagonal vertex coordinates of the second face detection frame.
  • the updating unit 504 may be further configured to: use the weight of the position information of the first face detection frame as the weight of the designated diagonal vertex coordinates of the first face detection frame, and use the second face detection frame as a weight.
  • the weight of the position information is used as the weight of the designated diagonal vertex coordinates of the second face detection frame, and the designated diagonal vertex coordinates of the first face detection frame and the designated diagonal vertex coordinates of the second face detection frame are used.
  • the weighted calculation result of is determined as the target diagonal vertex coordinates of the first face detection frame to update the position of the first face detection frame.
  • the specified diagonal vertex coordinates of the first face detection frame may include first vertex coordinates and second vertex coordinates, and the specified diagonal vertices of the second face detection frame.
  • the coordinates may include a third vertex coordinate and a fourth vertex coordinate.
  • the updating unit 504 may be further configured to: determine a weighted calculation result of the abscissa of the first vertex coordinate and the abscissa of the third vertex coordinate as the first target abscissa; and determine the ordinate of the first vertex coordinate
  • the weighted calculation result of the ordinate with the third vertex coordinate is determined as the first target ordinate
  • the weighted calculation result of the abscissa of the above second vertex coordinate and the abscissa of the fourth vertex coordinate is determined as the second target abscissa
  • the weighted calculation result of the vertical coordinate of the second vertex coordinate and the vertical coordinate of the fourth vertex coordinate is determined as the second target vertical coordinate
  • the second The coordinates formed by the target horizontal coordinate and the second target vertical coordinate are determined as the target diagonal vertex coordinates of the first face detection frame.
  • the device provided by the foregoing embodiment of the present application obtains, through the obtaining unit 501, the position information of the first face detection frame of the current frame of the target video and the position information of the second face detection frame of the previous frame. Therefore, the first determining unit 502 may determine an intersection ratio of the first face detection frame and the second face detection frame based on the obtained position information. After that, the second determining unit 503 determines the weight of the obtained position information of each face detection frame based on the intersection ratio, and can determine the weight of the obtained position information of each face detection frame. Finally, the updating unit 504 may determine the target position information of the first face detection frame based on the determined weight and the obtained position information to update the position of the first face detection frame.
  • the position of the face detection frame in the subsequent frames can be adjusted based on the intersection of the face detection frames in the two frames before and after. Because the position of the face detection frame in the subsequent frame considers the position of the face detection frame in the previous frame, and the entire area of the face detection frame in the previous frame is considered instead of a single coordinate, thereby reducing the The jitter of the face detection frame improves the smoothing effect and movement stability of the face detection frame in the video.
  • FIG. 6 illustrates a schematic structural diagram of a computer system 600 suitable for implementing an electronic device according to an embodiment of the present application.
  • the electronic device shown in FIG. 6 is only an example, and should not impose any limitation on the functions and scope of use of the embodiments of the present application.
  • the computer system 600 includes a central processing unit (CPU) 601, which can be loaded into a random access memory (RAM) 603 according to a program stored in a read-only memory (ROM) 602 or from a storage portion 608. Instead, perform various appropriate actions and processes.
  • RAM random access memory
  • ROM read-only memory
  • various programs and data required for the operation of the system 600 are also stored.
  • the CPU 601, the ROM 602, and the RAM 603 are connected to each other through a bus 604.
  • An input / output (I / O) interface 605 is also connected to the bus 604.
  • the following components are connected to the I / O interface 605: an input portion 606 including a keyboard, a mouse, and the like; an output portion 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), and the speaker; a storage portion 608 including a hard disk and the like; a communication section 609 including a network interface card such as a LAN card, a modem, and the like.
  • the communication section 609 performs communication processing via a network such as the Internet.
  • the driver 610 is also connected to the I / O interface 605 as necessary.
  • a removable medium 611 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as necessary, so that a computer program read therefrom is installed into the storage section 608 as necessary.
  • the process described above with reference to the flowchart may be implemented as a computer software program.
  • embodiments of the present disclosure include a computer program product including a computer program carried on a computer-readable medium, the computer program containing program code for performing a method shown in a flowchart.
  • the computer program may be downloaded and installed from a network through the communication portion 609, and / or installed from a removable medium 611.
  • CPU central processing unit
  • the computer-readable medium described in this application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the foregoing.
  • the computer-readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable Programming read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the foregoing.
  • a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in combination with an instruction execution system, apparatus, or device.
  • a computer-readable signal medium may include a data signal that is included in baseband or propagated as part of a carrier wave, and which carries computer-readable program code. Such a propagated data signal may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • the computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, and the computer-readable medium may send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device .
  • Program code embodied on a computer-readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
  • each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, which contains one or more functions to implement a specified logical function Executable instructions.
  • the functions noted in the blocks may also occur in a different order than those marked in the drawings. For example, two successively represented boxes may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved.
  • each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts can be implemented by a dedicated hardware-based system that performs the specified function or operation , Or it can be implemented with a combination of dedicated hardware and computer instructions.
  • the units described in the embodiments of the present application may be implemented by software or hardware.
  • the described unit may also be provided in a processor, for example, it may be described as: a processor includes an acquisition unit, a first determination unit, a second determination unit, and an update unit.
  • the names of these units do not constitute a limitation on the unit itself in some cases.
  • the update unit may also be described as a “unit that updates the position of the second face detection frame”.
  • the present application also provides a computer-readable medium, which may be included in the device described in the foregoing embodiments; or may exist alone without being assembled into the device.
  • the computer-readable medium carries one or more programs, and when the one or more programs are executed by the device, the device is caused to obtain a first human face obtained by performing face detection on a current frame of a target video in advance.
  • Position information of the detection frame and acquiring position information of a second face detection frame obtained by performing face detection on a previous frame of the current frame in advance; determining the first face detection based on the acquired position information The intersection ratio of the frame and the second face detection frame; based on the intersection ratio, determining the weight of the obtained position information of each face detection frame; based on the determined weight and the obtained position information, determining the first Target position information of a face detection frame to update the position of the first face detection frame.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Analysis (AREA)

Abstract

L'invention concerne un procédé et un dispositif de génération d'informations. Le procédé consiste : à acquérir des informations de position d'une première boîte de délimitation de visage obtenue par la réalisation d'une détection de visage sur une trame en cours d'une vidéo cible à l'avance, et à acquérir des informations de position d'une seconde boîte de délimitation de visage obtenue par la réalisation d'une détection de visage sur une trame antérieure à la trame en cours à l'avance (201) ; à déterminer l'intersection sur l'union de la première boîte de délimitation de visage et de la seconde boîte de délimitation de visage en fonction des informations de position obtenues (202) : en fonction de l'intersection sur l'union, à déterminer le poids des informations de position obtenues de chaque boîte de délimitation de visage (203) ; et en fonction du poids déterminé et des informations de position obtenues, à déterminer des informations de position cible de la première boîte de délimitation de visage et à mettre à jour la position de la première boîte de délimitation de visage (204). Le procédé de la présente invention permet d'améliorer l'effet de lissage d'une boîte de délimitation de visage.
PCT/CN2018/115974 2018-09-21 2018-11-16 Procédé et dispositif de génération d'informations WO2020056903A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811110674.1A CN109308469B (zh) 2018-09-21 2018-09-21 用于生成信息的方法和装置
CN201811110674.1 2018-09-21

Publications (1)

Publication Number Publication Date
WO2020056903A1 true WO2020056903A1 (fr) 2020-03-26

Family

ID=65224012

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/115974 WO2020056903A1 (fr) 2018-09-21 2018-11-16 Procédé et dispositif de génération d'informations

Country Status (2)

Country Link
CN (1) CN109308469B (fr)
WO (1) WO2020056903A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112990084A (zh) * 2021-04-07 2021-06-18 北京字跳网络技术有限公司 图像处理方法、装置和电子设备
WO2022237902A1 (fr) * 2021-05-14 2022-11-17 长沙智能驾驶研究院有限公司 Procédé, appareil, et dispositif de détection d'objet, et support de stockage informatique
CN115661798A (zh) * 2022-12-23 2023-01-31 小米汽车科技有限公司 确定目标区域的方法、装置、车辆及存储介质

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110677585A (zh) * 2019-09-30 2020-01-10 Oppo广东移动通信有限公司 目标检测框的输出方法、装置、终端及存储介质
CN113592874A (zh) * 2020-04-30 2021-11-02 杭州海康威视数字技术股份有限公司 图像显示方法、装置和计算机设备
CN112084992B (zh) * 2020-09-18 2021-04-13 北京中电兴发科技有限公司 一种人脸关键点检测模块中人脸框选取方法
CN112584089B (zh) * 2020-12-10 2021-05-07 浙江华创视讯科技有限公司 人脸亮度调节方法、装置、计算机设备和存储介质
CN112613462B (zh) * 2020-12-29 2022-09-23 安徽大学 一种加权交并比方法
CN113065457B (zh) * 2021-03-30 2024-05-17 广州繁星互娱信息科技有限公司 人脸检测点处理方法、装置、计算机设备及存储介质
CN113283349A (zh) * 2021-05-28 2021-08-20 中国公路工程咨询集团有限公司 基于目标锚框优选策略的交通基建施工目标监测系统与方法
CN113538274A (zh) * 2021-07-14 2021-10-22 Oppo广东移动通信有限公司 图像美颜处理方法、装置、存储介质与电子设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106952288A (zh) * 2017-03-31 2017-07-14 西北工业大学 基于卷积特征和全局搜索检测的长时遮挡鲁棒跟踪方法
CN107832741A (zh) * 2017-11-28 2018-03-23 北京小米移动软件有限公司 人脸特征点定位的方法、装置及计算机可读存储介质
CN108009473A (zh) * 2017-10-31 2018-05-08 深圳大学 基于目标行为属性视频结构化处理方法、系统及存储装置
CN108197568A (zh) * 2017-12-31 2018-06-22 广州二元科技有限公司 一种提升在数字影像中的人脸识别的处理方法
CN108196680A (zh) * 2018-01-25 2018-06-22 盛视科技股份有限公司 一种基于人体特征提取与检索的机器人视觉跟随方法

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9984282B2 (en) * 2015-12-10 2018-05-29 Perfect Corp. Systems and methods for distinguishing facial features for cosmetic application
CN106228112B (zh) * 2016-07-08 2019-10-29 深圳市优必选科技有限公司 人脸检测跟踪方法及机器人头部转动控制方法和机器人
CN106934817B (zh) * 2017-02-23 2020-11-10 中国科学院自动化研究所 基于多属性的多目标跟踪方法及装置
CN107330920B (zh) * 2017-06-28 2020-01-03 华中科技大学 一种基于深度学习的监控视频多目标追踪方法
CN108388879B (zh) * 2018-03-15 2022-04-15 斑马网络技术有限公司 目标的检测方法、装置和存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106952288A (zh) * 2017-03-31 2017-07-14 西北工业大学 基于卷积特征和全局搜索检测的长时遮挡鲁棒跟踪方法
CN108009473A (zh) * 2017-10-31 2018-05-08 深圳大学 基于目标行为属性视频结构化处理方法、系统及存储装置
CN107832741A (zh) * 2017-11-28 2018-03-23 北京小米移动软件有限公司 人脸特征点定位的方法、装置及计算机可读存储介质
CN108197568A (zh) * 2017-12-31 2018-06-22 广州二元科技有限公司 一种提升在数字影像中的人脸识别的处理方法
CN108196680A (zh) * 2018-01-25 2018-06-22 盛视科技股份有限公司 一种基于人体特征提取与检索的机器人视觉跟随方法

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112990084A (zh) * 2021-04-07 2021-06-18 北京字跳网络技术有限公司 图像处理方法、装置和电子设备
WO2022237902A1 (fr) * 2021-05-14 2022-11-17 长沙智能驾驶研究院有限公司 Procédé, appareil, et dispositif de détection d'objet, et support de stockage informatique
CN115661798A (zh) * 2022-12-23 2023-01-31 小米汽车科技有限公司 确定目标区域的方法、装置、车辆及存储介质

Also Published As

Publication number Publication date
CN109308469B (zh) 2019-12-10
CN109308469A (zh) 2019-02-05

Similar Documents

Publication Publication Date Title
WO2020056903A1 (fr) Procédé et dispositif de génération d'informations
CN105981368B (zh) 在成像装置中的照片构图和位置引导
EP3271865B1 (fr) Détection de segments d'un programme vidéo
BR112021005620A2 (pt) método e sistema para gerar conteúdo de mídia, e, mídia legível.
CN108197618B (zh) 用于生成人脸检测模型的方法和装置
US10192582B2 (en) Automatic generation of time-lapse videos
JP2017528016A (ja) ルールに基づくビデオ重要度解析
US20160054895A1 (en) Method of providing visual sound image and electronic device implementing the same
WO2020248900A1 (fr) Procédé et appareil de traitement de vidéo panoramique et support de stockage
US9729792B2 (en) Dynamic image selection
CN109271929B (zh) 检测方法和装置
CN104394422A (zh) 一种视频分割点获取方法及装置
KR20210040882A (ko) 동영상을 생성하기 위한 방법 및 장치
CN111436005B (zh) 用于显示图像的方法和装置
CN112954450B (zh) 视频处理方法、装置、电子设备和存储介质
CN110781823B (zh) 录屏检测方法、装置、可读介质及电子设备
WO2022007565A1 (fr) Procédé et appareil de traitement d'image pour réalité augmentée, dispositif électronique et support d'enregistrement
WO2020151156A1 (fr) Procédé et système d'évaluation de risque d'avant vente, appareil informatique et support de stockage lisible
WO2019062631A1 (fr) Procédé et dispositif de génération d'image dynamique locale
WO2018166470A1 (fr) Procédé d'affichage d'animation sur la base d'un débit de trames et dispositif terminal
CN109816791B (zh) 用于生成信息的方法和装置
US20210104096A1 (en) Surface geometry object model training and inference
CN111027495A (zh) 用于检测人体关键点的方法和装置
CN111314627B (zh) 用于处理视频帧的方法和装置
CN114546311A (zh) 用于智慧课堂的多显示终端投屏系统和方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18934390

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 08/07/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 18934390

Country of ref document: EP

Kind code of ref document: A1