BACKBONE NETWORK LEARNING METHOD AND SYSTEM BASED ON SELF-SUPERVISED LEARNING AND MULTI-HEAD FOR VISUAL INTELLIGENCE

    公开(公告)号:US20240394546A1

    公开(公告)日:2024-11-28

    申请号:US18225304

    申请日:2023-07-24

    Abstract: There is provided a learning method and system of a backbone network for visual intelligence based on self-supervised learning and multi-head. A network learning system according to an embodiment generates a plurality of first modified vectors by modifying a first feature vector outputted from a teacher network, generates a plurality of second modified vectors by modifying a second feature vector outputted from a student network, calculates a loss by using the first modified vectors and the second modified vectors, and optimizes parameters of the student network. Accordingly, the effect of learning by knowledge distillation may be enhanced by training the backbone network for visual intelligence like group learning is performed by various teacher networks and student networks.

    IMAGE INPAINTING APPARATUS AND IMAGE INPAINTING METHOD

    公开(公告)号:US20240378695A1

    公开(公告)日:2024-11-14

    申请号:US18032825

    申请日:2022-11-04

    Abstract: The present invention relates to an image inpainting apparatus and an image inpainting method, the image inpainting apparatus including: a background inpainting part configured to generate a background-inpainted image by carrying out inpainting on a background with respect to an input image in which a region to be inpainted is set up; an object inpainting part configured to generate an object image by carrying out inpainting on an object; and an image overlapping part configured to generate an output image by causing the background-inpainted image and the object image, which are generated, to overlap each other.

    MICROPLASTIC DETECTION SENSOR AND MICROPLASTIC DETECTION SYSTEM USING THE SAME

    公开(公告)号:US20240337575A1

    公开(公告)日:2024-10-10

    申请号:US18405560

    申请日:2024-01-05

    CPC classification number: G01N15/0272 B01D29/44 G01N15/06 G01N2015/0053

    Abstract: Proposed is a microplastic detection sensor for detecting information about microplastic contained in a sample. The sensor may include a fluidic channel substrate including an inlet and an outlet, and a plurality of RF resonance structures. The inlet may be formed on one end of the fluidic channel substrate, and the outlet may be formed on the other end. The fluidic channel substrate may have a microfluidic channel formed therein to connect the inlet and the outlet. The microfluidic channel may move the sample toward the outlet by capillary action. In the fluidic channel substrate, a plurality of capture parts, respectively corresponding to the RF resonance structures, may be formed along the microfluidic channel and may selectively capture the microplastic by particle size. The RF resonance structures may output information about the microplastic captured in the corresponding capture part through RF resonance for the applied RF signal.

    METHOD AND SYSTEM FOR SELECTING OPTIMAL SYNCHRONIZATION BROKER FOR EACH WORKLOAD

    公开(公告)号:US20240236177A9

    公开(公告)日:2024-07-11

    申请号:US18380968

    申请日:2023-10-17

    CPC classification number: H04L67/1008 H04L67/1012

    Abstract: There are provided a method and a system for selecting an optimal synchronization broker for each workload. A system for selecting an optimal synchronization broker for each workload according to an embodiment includes: a storage unit including a database that contains specialist recommendation information including information on a correct protocol that is recommended by a specialist with respect to a service specification and data transmission requirements; and a processor configured to determine an optimal in-bound protocol and an optimal out-bound protocol according to the data transmission requirements through a decision tree model that is trained by using the specialist recommendation information as training data.

Patent Agency Ranking