SYSTEMS, METHODS, AND APPARATUSES FOR IMPLEMENTING TRANSPARENT MODELS FOR COMPUTER VISION AND IMAGE RECOGNITION UTILIZING DEEP LEARNING NON-TRANSPARENT BLACK BOX MODELS

    公开(公告)号:US20250095347A1

    公开(公告)日:2025-03-20

    申请号:US18293315

    申请日:2022-08-24

    Inventor: Roy Asim

    Abstract: Described herein are means for systematically generating transparent models for computer vision and image recognition utilizing deep learning non-transparent black box models. According to a particular embodiment, there is a specially configured system for generating an explainable AI model by performing operations, including: training a Convolutional Neural Network (CNN) to classify objects; training a Convolutional Neural Network (CNN) to classify objects from training data having a set of training images; training a multi-layer perceptron (MLP) to recognize both the objects and parts of the objects; generating the explainable AI model based on the training of the MLP; receiving an image having an object embedded therein, wherein the image forms no portion of the training data for the explainable AI model; executing the CNN and the explainable AI model within an image recognition system, and generating a prediction of the object in the image via the explainable AI model; recognizing parts of the object; providing the parts recognized within the object as evidence for the prediction of the object; and generating a description of why the image recognition system predicted the object in the image based on the evidence comprising the recognized parts.

    OCCLUSION-CAPABLE OPTICAL VIEWING DEVICE AND ASSOCIATED METHOD

    公开(公告)号:US20250093654A1

    公开(公告)日:2025-03-20

    申请号:US18726715

    申请日:2023-01-07

    Abstract: An occlusion-capable viewing device includes an imaging lens, a collimating lens, and a virtual display module, and between the imaging lens and the collimating lens, an image inverter, and a spatial light modulator. The imaging lens projects a first inverted image of a scene on to an entrance surface of the image inverter. The image inverter rotates the first inverted image to yield a first upright image. The spatial light modulator attenuates a first image-region of the first upright image to produce a modulated optical beam. The virtual display module includes a display device and a light combiner, which combines (i) the collimated optical beam and (ii) illumination emitted by the display device to yield an occluded image.

    DIFFRACTIVE TRIFOCAL LENS
    27.
    发明申请

    公开(公告)号:US20250076683A1

    公开(公告)日:2025-03-06

    申请号:US18954376

    申请日:2024-11-20

    Abstract: A diffractive multifocal lens is disclosed, comprising an optical element having at least one diffractive surface, the surface profile comprising a plurality of annular concentric zones. The optical thickness of the surface profile changes monotonically with radius within each zone, while a distinct step in optical thickness at the junction between adjacent zones defines a step height. The step heights for respective zones may differ from one zone to another periodically so as to tailor diffraction order efficiencies of the optical element. In one example of a trifocal lens, step heights alternate between two values, the even-numbered step heights being lower than the odd-numbered step heights. By plotting a topographical representation of the diffraction efficiencies resulting from such a surface profile, step heights may be optimized to direct a desired level of light power into the diffraction orders corresponding to near, intermediate, and distance vision, thereby optimizing the lens performance.

Patent Agency Ranking