Method and system for meaningful counterfactual explanations

    公开(公告)号:US11961287B2

    公开(公告)日:2024-04-16

    申请号:US17449872

    申请日:2021-10-04

    摘要: A computer-implemented method for explaining an image classifier, the method comprising: receiving an initial image, the initial image having been wrongly classified by the image classifier; receiving an initial gradient of a function executed by the image classifier generated while classifying the initial image, the function being indicative of a probability for the initial image to belong to an initial class; converting the initial image into a latent vector, the latent vector being a representation of the initial image in a latent space; generating a plurality of perturbation vectors using the initial gradient of the function executed by the image classifier; combining the latent vector with each one of the plurality of perturbation vectors, thereby obtaining a plurality of modified vectors; for each one of the plurality of modified vectors, reconstructing a respective image, thereby obtaining a plurality of reconstructed images; transmitting the reconstructed images to the image classifier; for each one of the plurality of reconstructed images, receiving a respective updated gradient of the function executed by the image classifier; using the respective updated gradients, determining amongst the reconstructed images at least one given reconstructed image for which the respective updated gradient is indicative that a new class different from the initial class has been assigned by the image classifier; and outputting the at least one given reconstructed image.

    Bus bar for wind turbine
    2.
    外观设计

    公开(公告)号:USD1001734S1

    公开(公告)日:2023-10-17

    申请号:US29762325

    申请日:2020-12-16

    摘要: FIG. 1 is a perspective view of a bus bar for wind turbine, showing our design;
    FIG. 2 is a front view thereof;
    FIG. 3 is a rear view thereof;
    FIG. 4 is a left view thereof;
    FIG. 5 is a right view thereof;
    FIG. 6 is a top view thereof; and,
    FIG. 7 is a bottom view thereof.
    The bus bar for wind turbine is shown with a symbolic break in its length and height as illustrated with the dash-dot broken lines. The appearance of any portion of the bus bar for wind turbine between the break lines forms no part of the claimed design. In the drawings, the evenly spaced broken lines illustrate portions of the bus bar for wind turbine which form no part of the claimed design.

    Wheel chock
    4.
    外观设计

    公开(公告)号:USD987542S1

    公开(公告)日:2023-05-30

    申请号:US29775247

    申请日:2021-03-22

    摘要: FIG. 1 is a front right isometric view of a wheel chock showing our new design;
    FIG. 2 is a rear left isometric view thereof;
    FIG. 3 is a right side view thereof;
    FIG. 4 is a left side view thereof;
    FIG. 5 is a front view thereof;
    FIG. 6 is a rear view thereof;
    FIG. 7 is a top view thereof; and,
    FIG. 8 is a bottom view thereof.
    The wheel chock illustrated in FIGS. 1-8 is a first embodiment. Our new design also includes a second embodiment which is mirror symmetrical to FIGS. 1-8 and therefore not separately illustrated.

    Method and system for generating a vector representation of an image

    公开(公告)号:US11645505B2

    公开(公告)日:2023-05-09

    申请号:US16746157

    申请日:2020-01-17

    申请人: ELEMENT AI Inc.

    IPC分类号: G06K9/62 G06N3/08

    摘要: There is described a computer-implemented method for generating a vector representation of an image, the computer-implemented method comprising: receiving a given image and semantic information about the given image; generating a first vector representation of the given image using an image embedding method; generating a second vector representation of the semantic information using a word embedding method; combining the first vector representation of the image to be embedded and the second vector representation of the semantic information together, thereby obtaining a modified vector representation for the image to be embedded; and outputting the modified vector representation.

    Dual channel medical simulator
    7.
    发明授权

    公开(公告)号:US11495141B2

    公开(公告)日:2022-11-08

    申请号:US15940402

    申请日:2018-03-29

    摘要: An apparatus for simulating an insertion of an elongated instrument into a subject, comprising: a frame extending between two end walls along a first axis and two lateral walls along a second axis, one of the two end walls being provided with an insertion aperture and one of the two lateral walls being provided an insertion hole, the insertion aperture defining a first passageway and the insertion hole defining a second passageway, the first and second passageways intersecting each other at an intersection point; and a sensing unit contained within the frame and configured for measuring at least one of a displacement of the elongated member and a rotation of the elongated member, the sensing unit being positioned adjacent to the intersection point for performing the measurement of the at least one of the displacement and the rotation at the intersection point.

    Valve assembly for a gas chromatograph

    公开(公告)号:US11448623B2

    公开(公告)日:2022-09-20

    申请号:US16914292

    申请日:2020-06-27

    申请人: LDETEK INC.

    摘要: The valve assembly includes a main valve body defining a first inner cylindrical cavity, a support plate positioned above the main valve body, and an upper valve head positioned above the support plate. The upper valve head includes a bottom planar bottom surface and a plurality of discrete conduits, each conduit extending between an outer end and an inner end. The valve assembly further includes a sealing diaphragm interposed between the support plate and the upper valve head. The diaphragm can be engaged by plungers to selectively open or close corresponding gas circuits.

    Artificial eye system
    9.
    发明授权

    公开(公告)号:US11263924B2

    公开(公告)日:2022-03-01

    申请号:US15475730

    申请日:2017-03-31

    IPC分类号: G09B23/32 G09G3/34 G09G3/32

    摘要: An artificial eye system comprising: a display unit for displaying an image of an eye thereon, the display unit comprising a light cell adapted to selectively emit light and detect light; and a controller for controlling the display unit, the controller for: operating the display unit to concurrently detect light through the light cell and display an actual image of the eye; determining a next mage of the eye to be displayed on the display unit as a function of the detected light; and operating the display unit to display the next image.

    System and method for preparing crustacean body parts for meat extraction

    公开(公告)号:US10765123B2

    公开(公告)日:2020-09-08

    申请号:US15995857

    申请日:2018-06-01

    摘要: A system and method for preparing a shelled crustacean body part for meat extraction are provided. The method involves capturing an image of said body part, evaluating the thickness of the shell from the image, determining from the thickness evaluation at least one laser parameter for forming a line of weakness within the shell, the line of weakness being dimensioned to facilitate meat extraction, while preserving the sensory characteristics of the meat, and sweeping a laser beam on the body part in accordance with the laser parameter to form the line of weakness within the shell. The method and system provide the advantage of facilitating meat extraction without compromising the quality or commercial value of the meat all while increasing productivity and performance in comparison with completely manual extraction techniques.