KNOTTER DEVICE
    3.
    发明公开
    KNOTTER DEVICE 审中-公开

    公开(公告)号:EP4501831A1

    公开(公告)日:2025-02-05

    申请号:EP23779071.2

    申请日:2023-02-22

    Abstract: There is provided a knotter device capable of tying a sheet bend knot stably.
    A yarn joining unit 20 includes a yarn guide 23 configured to guide an old yarn 2b in a direction different from the direction in which the old yarn 2b extends; a knitting needle 21 configured to hook the old yarn 2b and to twist the old yarn 2b to form a twisted stitch 4, to hook another part of the old yarn 2b guided by the yarn guide 23, to form a stitch 5 by passing the other part through the twisted stitch 4 to one side, and to pass the new yarn 2a guided by the yarn guide lever 14 through the stitch 5; a hook 22 around which the old yarn 2b is wound when the knitting needle 21 forms the stitch 5 and that is operable to unwind the old yarn, and a tensioner 27 and a yarn guide lever 14 configured to increase the tension of the old yarn 2b so that the stitch 5 having the new yarn 2a passed is passed through the twisted stitch 4, to the other side.

    STORAGE AUXILIARY DEVICE, STORAGE AUXILIARY METHOD, AND PROGRAM

    公开(公告)号:EP4485230A1

    公开(公告)日:2025-01-01

    申请号:EP23774530.2

    申请日:2023-03-08

    Abstract: A data acquiring unit (21) that acquires visual information or auditory information, and that acquires brain activation information; a tag generating unit (32) that generates a tag in which the visual information or the auditory information, the brain activation information, and a keyword are associated with an object identified from the visual information; and a storage unit (14) that stores the tag.

    IMAGE CLASSIFYING DEVICE, IMAGE CLASSIFYING METHOD, AND IMAGE CLASSIFYING PROGRAM, AND IMAGE FEATURE LEARNING DEVICE, IMAGE FEATURE LEARNING METHOD, AND IMAGE FEATURE LEARNING PROGRAM

    公开(公告)号:EP4471711A1

    公开(公告)日:2024-12-04

    申请号:EP23747028.1

    申请日:2023-01-26

    Inventor: TAKEHARA Hideki

    Abstract: A feature extraction unit (210) extracts a low-resolution general feature vector of an input image and a high-resolution detailed feature vector of the input image. A general feature distance measurement unit (222a) maintains a general weight vector of each class and calculates a general distance vector from the general feature vector and the general weight vector. A detailed feature distance measurement unit (222b) maintains a detailed weight vector of each class and calculates a detailed distance vector from the detailed feature vector and the detailed weight vector. A feature distance synthesis unit (230) calculates a synthesized distance vector by synthesizing the general distance vector and the detailed distance vector. A global classification unit (260) determines a class of the input image based on the synthesized distance vector.

Patent Agency Ranking