公开(公告)号：US11805860B2

公开(公告)日：2023-11-07

申请号：US18064029

申请日：2022-12-09

Applicant: ASICS CORPORATION

Inventor： Hiroyuki Kusumi ,  Toshiaki Okamoto ,  Genki Hatano ,  Masaru Ichikawa

IPC: A43D1/02

CPC classification number: A43D1/02

Abstract: A prediction device includes: a communication device that acquires measurement subject data including measurement data of a shape of a foot of a measurement subject person in a loaded state; a storage that stores first sample data in the loaded state and second sample data in an unloaded state, the first sample data and the second sample data being calculated from measurement data of foot shapes of a plurality of samples, the samples being identical both in the loaded state and the unloaded state; and a processor that predicts the shape of the foot of the measurement subject person in the unloaded state. The processor calculates a difference between the measurement subject data and the first sample data, and predicts the shape of the foot of the measurement subject person in the unloaded state based on the difference and the second sample data.

公开(公告)号：US11647814B2

公开(公告)日：2023-05-16

申请号：US17255967

申请日：2019-06-26

Applicant: ASICS CORPORATION

Inventor： Masaru Ichikawa ,  Hiroyuki Kusumi ,  Kazunari Takeichi

IPC: A43D1/02 ,  G06N20/00 ,  A43D1/06

CPC classification number: A43D1/02 ,  A43D1/06 ,  G06N20/00

Abstract: In a markerless foot size estimation device 100, a shape data input unit 10 acquires three-dimensional shape data of a foot of a subject and stores the three-dimensional shape data in a shape data storage unit 20. A characteristic extraction unit 30 extracts foot shape characteristics from the three-dimensional shape data of a subject's foot stored in the shape data storage unit 20, and stores the foot shape characteristics in a shape characteristic storage unit 40. A machine learning unit 50 uses, as teacher data, foot shape characteristics and arch height stored in the shape characteristic storage unit 40 to create, through machine learning, an estimation model used to estimate arch height based on foot shape characteristics, and stores the estimation model in an estimation model storage unit 60. An estimation output unit 70 uses the estimation model stored in the estimation model storage unit 60 to estimate arch height based on extracted shape characteristics and outputs the arch height.