TARGET RE-IDENTIFICATION METHOD, NETWORK TRAINING METHOD THEREOF, AND RELATED DEVICE

    公开(公告)号:WO2022001034A1

    公开(公告)日:2022-01-06

    申请号:PCT/CN2020/139349

    申请日:2020-12-25

    Abstract: A target re-identification method, a network training method thereof and a related device. The training method comprises the steps of obtaining a training image set; identifying each training image in the training image set by using a target re-identification network; obtaining an identification result of each training image; wherein the target re-identification network comprises a plurality of branches; wherein the recognition result of each training image comprises feature information output by each branch and a classification result corresponding to the feature information; wherein the feature information output by one branch comprises n pieces of local feature information, n is greater than 3, and the n pieces of local feature information correspond to different image areas of the training image; and adjusting the parameter of each branch of the target re-identification network based on the identification result of the training image. In this way, the result of target recognition by the trained target re-recognition network is more accurate.

    AFFINE PREDICATION METHODANDCOMPUTING DEVICE

    公开(公告)号:WO2021052478A1

    公开(公告)日:2021-03-25

    申请号:PCT/CN2020/116275

    申请日:2020-09-18

    Abstract: An affine prediction method anda computing device are disclosed. The method includes obtaining a current coding block comprising a plurality of sub-blocks arranged in an array; dividing the current coding block into at least two areas, wherein each area of the at least two areas comprises two or more of the sub-blocks; setting a corresponding control point motion vector (CPMV) for one or more control points in the each area, respectively; and affine-predicting motion vectors (MVs) of the two or more of the sub-blocks in the each area, using the corresponding CPMV.

    SYSTEMS AND METHODS FOR VIDEO ENCODING
    5.
    发明申请

    公开(公告)号:WO2022117113A1

    公开(公告)日:2022-06-09

    申请号:PCT/CN2021/135831

    申请日:2021-12-06

    Abstract: The present disclosure relates to systems and methods for video encoding. The systems may determine a current string in a coding unit (CU) of an image frame. The current string may include one or more pixels. The systems may determine, based on a first pixel in the current string, a reference matching pixel of the first pixel by searching a searching area. The searching area may include a plurality of matching pixels The systems may determine, based on the reference matching pixel and matching patterns, a count of matching pixels in each of a plurality of groups, corresponding to the matching patterns, of the current string. The systems may further determine, based on counts of matching pixels in the plurality of groups corresponding to the matching patterns, a target prediction of the current string.

    SYSTEMS AND METHODS FOR VIDEO PROCESSING
    6.
    发明申请

    公开(公告)号:WO2022117104A1

    公开(公告)日:2022-06-09

    申请号:PCT/CN2021/135660

    申请日:2021-12-06

    Abstract: The present disclosure relates to systems and methods for video encoding. The systems may determine a current string in a coding unit (CU) of an image frame. The current string may include one or more pixels. The systems may determine a first cost based on pixels with a same pixel value in the current string. The systems may determine a second cost based on one or more strings that are located above the current string. The systems may further determine a prediction string of the current string based on the first cost and the second cost. To determine the prediction string of the current string based on the first cost and the second cost, the systems may designate the current string as the prediction string of the current string in response to that the first cost and the second cost are not determined.

    METHODS AND SYSTEMS FOR INTRA-PREDICTION
    7.
    发明申请

    公开(公告)号:WO2021115471A1

    公开(公告)日:2021-06-17

    申请号:PCT/CN2020/136034

    申请日:2020-12-14

    Abstract: A method for intra-prediction is provided, including: determining at least one reference line of a current coding block, each of the at least one reference line comprising a plurality of first reference pixels; down-sampling the at least one reference line and obtaining a plurality of second reference pixels; performing matrix multiplication for the plurality of second reference pixels and obtaining a reference matrix; and obtaining a matrix-based intra-prediction block of the current coding block based on the reference matrix and the plurality of first reference pixels, the matrix-based intra-prediction block comprising a plurality of first pixels and a plurality of second pixels, a value of each of the first pixels is a value of a corresponding pixel in the reference matrix, and a value of each of the second pixels is obtained based on values of the first pixels and values of the first reference pixels.

    SYSTEMS AND METHODS FOR INTER-FRAME PREDICTION

    公开(公告)号:WO2020259610A1

    公开(公告)日:2020-12-30

    申请号:PCT/CN2020/098182

    申请日:2020-06-24

    Abstract: The present disclosure provides systems and methods for inter-frame prediction. A method may include obtaining a current coding unit in an image. The method may include obtaining a first MV candidate list corresponding to the current coding unit. The first MV candidate list may include at least one of a first bidirectional MV or a first unidirectional MV. The method may also include determining, based on the first MV candidate list, a second MV candidate list corresponding to the current coding unit and determining, based on the second MV candidate list, a target prediction unit of the current coding unit.

    SYSTEMS AND METHODS FOR IMAGE CODING
    9.
    发明申请

    公开(公告)号:WO2020259263A1

    公开(公告)日:2020-12-30

    申请号:PCT/CN2020/094832

    申请日:2020-06-08

    Abstract: The present disclosure relates to systems and methods for image coding. The system may determine a coding unit in an image and divide the coding unit into a plurality of coding sub-units. For at least one coding sub-unit of the plurality of coding sub-units, the system may determine multiple reference lines of the at least one coding sub-unit. The multiple reference lines may include a first reference line closest to the at least one coding sub-unit and one or more second reference lines. The system may determine one or more most probable modes (MPM) lists associated with at least one coding sub-unit. The one or more MPM lists may include at least one first MPM list corresponding to the first reference line and at least one second MPM list corresponding to the one or more second reference lines. The system may determine a target prediction mode associated with one of the at least one coding sub-unit based on the one or more MPM lists. The system may determine a predicted coding sub-unit corresponding to the one of the at least one coding sub-unit based on the target prediction mode. The system may determine a predicted coding unit corresponding to the coding unit based at least in part on the predicted coding sub-unit.

    UPDATING METHOD FOR SIMILARITY THRESHOLD IN FACE RECOGNITION AND ELECTRONIC DEVICE

    公开(公告)号:WO2023284185A1

    公开(公告)日:2023-01-19

    申请号:PCT/CN2021/128816

    申请日:2021-11-04

    Abstract: An updating method for a similarity threshold in face recognition is disclosed. The updating method includes receiving a first face image and obtaining at least one attribute of the first face image by performing attribute analysis on the first face image; obtaining a plurality of similarity values by performing feature comparison on the first face image and each of a plurality of second face images in a database; adding the plurality of similarity values into at least one attribute error report based on the at least one attribute of the first face image; obtaining a candidate similarity value from any one of the at least one attribute error report, in response to the any one of the at least one attribute error report having the number of similarity values which is greater than or equal to a first threshold; and updating a current similarity threshold through the candidate similarity value.

Patent Agency Ranking