利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

274 条记录 (耗时 0.014 秒)

    MOTION REPRESENTATION CALCULATION METHOD AND SYSTEM, TRAINING METHOD, COMPUTER PROGRAM, READABLE MEDIUM AND SYSTEM
    73.
    发明公开
    MOTION REPRESENTATION CALCULATION METHOD AND SYSTEM, TRAINING METHOD, COMPUTER PROGRAM, READABLE MEDIUM AND SYSTEM 审中-公开

    公开(公告)号：EP4300435A1

    公开(公告)日：2024-01-03

    申请号：EP22305979.1

    申请日：2022-07-01

    申请人： TOYOTA JIDOSHA KABUSHIKI KAISHA ,  Institut National de Recherche en Informatique et en Automatique (INRIA)

    发明人： FRANCESCA, Gianpiero ,  GARATTONI, Lorenzo ,  YANG, Di ,  WANG, Yaohui ,  BREMOND, François

    IPC分类号： G06V10/82 ,  G06V10/774 ,  G06V20/40

    摘要： A method for training a motion encoder (ME) and/or a character encoder (CE) to calculate a motion or a character representation (r m ) representing a motion or a character of a person, based on a skeleton sequence (p m,c ), comprising: S11)-S12) based on two initial skeleton sequence (p m',c' ), calculating respectively a first and second primary motion representation (r1 m ) with the motion encoder (ME) and a first and second primary character representation (r1 c ) with the character encoder (CE);
    S21)-S22) based on the first and second primary motion representations (r1 m , r2 m ) and the first and second primary character representations (r1 c' ,r2 c' ), calculating a first and a second primary retargeted skeleton sequence (p1 m,c' ,p1 m',c );
    S31)-S32) based on the first and second primary skeleton sequence (p1 m,c' , p1 m',c ), calculating a first and second secondary motion representation (r2 m ) with the motion encoder (ME) and calculating a first and second secondary character representation (r2 c' ) with the character encoder (CE);
    S41)-S42) based on the first and second secondary motion representation (r2 m , r2 m' ) and the second secondary character representation (r2 c , r2 c' ), calculating a first and second secondary retargeted skeleton sequence (p2 m,c );
    S60) updating parameters of the motion encoder (ME) and/or the character encoder (CE).

    METHOD FOR TRAINING AN OBJECT RECOGNITION MODEL IN A COMPUTING DEVICE
    77.
    发明公开
    METHOD FOR TRAINING AN OBJECT RECOGNITION MODEL IN A COMPUTING DEVICE 审中-公开

    公开(公告)号：EP4283529A1

    公开(公告)日：2023-11-29

    申请号：EP23151976.0

    申请日：2023-01-17

    申请人： Hanwha Vision Co., Ltd.

    发明人： Park, Sungyeon ,  Shin, Hyunhak ,  Song, Changho ,  Noh, Seungin ,  Lim, Jeongeun

    IPC分类号： G06N3/084 ,  G06N3/096 ,  G06N3/0464 ,  G06V10/20 ,  G06V10/774 ,  G06V10/82 ,  G06V10/98 ,  G06V20/52

    摘要： An object recognition model training method in a computing device is disclosed. In the present disclosure, an object of interest, which is an object for object recognition, is designated, and an object of non-interest excluding the object of interest is generated and used as learning data for the object recognition model. In the process of training the object recognition model, when an erroneously detected object occurs, the object recognition model may be retrained by automatically converting the erroneously detected object to the object of non-interest without feedback of the erroneous detection to the user. Accordingly, user convenience for processing the erroneously detected object is improved, which increases reliability of the object recognition model. This disclosure can be associated with artificial intelligence modules, drones (unmanned aerial vehicles (UAVs)), robots, augmented reality (AR) devices, virtual reality (VR) devices, devices related to 5G service, etc.