公开(公告)号：US11669945B2

公开(公告)日：2023-06-06

申请号：US16858862

申请日：2020-04-27

Applicant: GE Precision Healthcare LLC

Inventor： Tao Tan ,  Pál Tegzes ,  Levente Imre Török ,  Lehel Ferenczi ,  Gopal B. Avinash ,  László Ruskó ,  Gireesha Chinthamani Rao ,  Khaled Younis ,  Soumya Ghose

IPC: G06F18/21 ,  G06V10/772 ,  G06V10/774 ,  G06V10/762 ,  G06V10/74 ,  G06V10/776 ,  G06T7/00 ,  G06V10/82 ,  G06F18/22 ,  G06F18/23 ,  G06F18/28 ,  G06F18/214

CPC classification number: G06F18/217 ,  G06F18/214 ,  G06F18/22 ,  G06F18/23 ,  G06F18/28 ,  G06T7/00 ,  G06V10/761 ,  G06V10/762 ,  G06V10/772 ,  G06V10/774 ,  G06V10/776 ,  G06V10/82

Abstract: Techniques are described for optimizing deep learning model performance using image harmonization as a pre-processing step. According to an embodiment, a method comprises decomposing, by a system operatively coupled to a processor, an input image into sub-images. The method further comprises harmonizing the sub-images with corresponding reference sub-images of at least one reference image based on two or more different statistical values respectively calculated for the sub-images and the corresponding reference-sub images, resulting in transformation of the sub-images into modified sub-images images. In some implementations, the modified sub-images can be combined into a harmonized image having a more similar appearance to the at least one reference image relative to the input image. In other implementations, harmonized images and/or modified sub-images generated using these techniques can be used as ground-truth training samples for training one or more deep learning model to transform input images with appearance variations into harmonized images.

公开(公告)号：US12125231B2

公开(公告)日：2024-10-22

申请号：US17579716

申请日：2022-01-20

Applicant: Mobileye Vision Technologies Ltd.

Inventor： Ishay Golinsky ,  Iddo Hanniel ,  Moshe Shenfeld ,  Yael Hacohen

IPC: G06T7/70 ,  G01C21/00 ,  G01C21/16 ,  G01C21/32 ,  G01C21/36 ,  G05D1/00 ,  G05D1/249 ,  G06V10/772 ,  G06V20/56 ,  G06V20/58 ,  G08G1/16

CPC classification number: G06T7/70 ,  G01C21/005 ,  G01C21/16 ,  G01C21/32 ,  G01C21/3658 ,  G01C21/3837 ,  G01C21/3844 ,  G05D1/0246 ,  G05D1/249 ,  G06V10/772 ,  G06V20/584 ,  G06V20/588 ,  G08G1/167 ,  G06T2207/30244 ,  G06T2207/30261

Abstract: Systems and methods are provided for navigating a host vehicle. In an embodiment, a processing device may be configured to receive at least one image captured by an image capture device, the at least one image being representative of an enviromnent of the host vehicle; analyze the at least one image to identify an object in the environment of the host vehicle; determine a location of the host vehicle; receive map information associated with the determined location of the host vehicle, wherein the map information includes lane width information associated with a road in the environment of the host vehicle; determine a distance from the host vehicle to the object based on at least the lane width information; and determine a navigational action for the host vehicle based on the determined distance.

公开(公告)号：US12106554B2

公开(公告)日：2024-10-01

申请号：US17619968

申请日：2019-06-18

Applicant: XZIMG LIMITED

Inventor： Nicolas Livet

IPC: G06V10/82 ,  G06N3/044 ,  G06N3/084 ,  G06T7/20 ,  G06V10/26 ,  G06V10/764 ,  G06V10/77 ,  G06V10/772 ,  G06V10/774 ,  G06V10/80

CPC classification number: G06V10/82 ,  G06N3/044 ,  G06N3/084 ,  G06T7/20 ,  G06V10/26 ,  G06V10/764 ,  G06V10/7715 ,  G06V10/772 ,  G06V10/774 ,  G06V10/80 ,  G06T2207/10016 ,  G06T2207/20084

Abstract: A recurrent multi-task CNN with an encoder and multiple decoders infers single value output and dense (image) outputs such as heatmaps and segmentation masks. Recurrence is obtained by reinjecting (with mere concatenation) heatmaps or masks (or intermediate feature maps) to a next input image (or to next intermediate feature maps) for a next CNN inference. The inference outputs may be refined using cascaded refiner blocks specifically trained. Virtual annotation for training video sequences can be obtained using computer analysis. Benefits of these approaches allows the depth of the CNN, i.e. the number of layers, to be reduced. They also avoid parallel independent inferences to be run for different tasks, while keeping similar prediction quality. Multiple task inferences are useful for Augmented Reality applications.

公开(公告)号：US12093304B2

公开(公告)日：2024-09-17

申请号：US17279803

申请日：2019-08-19

Applicant: AMGEN INC.

Inventor： Killian Ryan

IPC: G06F16/50 ,  G06F16/51 ,  G06F18/214 ,  G06F18/28 ,  G06T1/00 ,  G06T7/00 ,  G06V10/772 ,  H04N1/00 ,  H04N5/00

CPC classification number: G06F16/50 ,  G06F16/51 ,  G06F18/2148 ,  G06F18/28 ,  G06T1/0007 ,  G06T7/0004 ,  G06V10/772 ,  H04N1/00 ,  H04N1/0009 ,  H04N5/00 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06V2201/06

Abstract: A method for sampling images includes receiving a first image set generated by automated imaging equipment during a first inspection period, and storing in a memory an image library that initially consists of the first image set. A plurality of new image sets is then sequentially received (302) during respective inspection periods. While the new image sets are received (302), the image library stored in the memory is updated. Updating the image library includes, for each new image set, adding to the image library a certain number of images distributed among the new image set and removing from the image library the same number of images distributed among a current instance of the image library (308). The number of overwritten images in the image library decreases from one inspection period to the next.

公开(公告)号：US12062225B2

公开(公告)日：2024-08-13

申请号：US17615946

申请日：2020-05-25

Applicant: KONINKLIJKE PHILIPS N.V.

Inventor： Bart Jacob Bakker ,  Dimitrios Mavroeidis ,  Stojan Trajanovski

IPC: G06V10/764 ,  G06V10/772 ,  G06V10/774 ,  G06V10/82

CPC classification number: G06V10/764 ,  G06V10/772 ,  G06V10/774 ,  G06V10/82

Abstract: Aspects and embodiments relate to a method of providing a representation of a feature identified by a deep neural network as being relevant to an outcome, a computer program product and apparatus configured to perform that method. The method comprises: providing the deep neural network with a training library comprising: a plurality of samples associated with the outcome; using the deep neural network to recognise a feature in the plurality of samples associated with the outcome; creating a feature recognition library from an input library by identifying one or more elements in each of a plurality of samples in the input library which trigger recognition of the feature by the deep neural network; using the feature recognition library to synthesise a plurality of one or more elements of a sample which have characteristics which trigger recognition of the feature by the deep neural network; and using the synthesised plurality of one or more elements to provide a representation of the feature identified by the deep neural network in the plurality of samples associated with the outcome. Accordingly, rather than visualising a single instance of one or more elements in a sample which trigger a feature associated with an outcome, it is possible to visualise a range of samples including elements which would trigger a feature associated with an outcome, thus enabling a more comprehensive view of operation of a deep neural network in relation to a particular feature.

公开(公告)号：US12051178B2

公开(公告)日：2024-07-30

申请号：US18304947

申请日：2023-04-21

Applicant: GE Precision Healthcare LLC

Inventor： Tao Tan ,  Pál Tegzes ,  Levente Imre Török ,  Lehel Ferenczi ,  Gopal B. Avinash ,  László Ruskó ,  Gireesha Chinthamani Rao ,  Khaled Younis ,  Soumya Ghose

IPC: G06T5/50 ,  G06F18/21 ,  G06F18/214 ,  G06F18/22 ,  G06F18/23 ,  G06F18/28 ,  G06T7/00 ,  G06V10/74 ,  G06V10/762 ,  G06V10/772 ,  G06V10/774 ,  G06V10/776 ,  G06V10/82

CPC classification number: G06T5/50 ,  G06F18/214 ,  G06F18/217 ,  G06F18/22 ,  G06F18/23 ,  G06F18/28 ,  G06T7/00 ,  G06V10/761 ,  G06V10/762 ,  G06V10/772 ,  G06V10/774 ,  G06V10/776 ,  G06V10/82

Abstract: Techniques are described for optimizing deep learning model performance using image harmonization as a pre-processing step. According to an embodiment, a method comprises decomposing, by a system operatively coupled to a processor, an input image into sub-images. The method further comprises harmonizing the sub-images with corresponding reference sub-images of at least one reference image based on two or more different statistical values respectively calculated for the sub-images and the corresponding reference-sub images, resulting in transformation of the sub-images into modified sub-images images. In some implementations, the modified sub-images can be combined into a harmonized image having a more similar appearance to the at least one reference image relative to the input image. In other implementations, harmonized images and/or modified sub-images generated using these techniques can be used as ground-truth training samples for training one or more deep learning model to transform input images with appearance variations into harmonized images.

公开(公告)号：US20240233177A9

公开(公告)日：2024-07-11

申请号：US18208660

申请日：2023-06-12

Applicant: Cognex Corporation

Inventor： Nathaniel R. Bogan

IPC: G06T7/73 ,  G06F18/28 ,  G06T7/00 ,  G06T7/33 ,  G06V10/44 ,  G06V10/46 ,  G06V10/75 ,  G06V10/772

CPC classification number: G06T7/75 ,  G06F18/28 ,  G06T7/001 ,  G06T7/344 ,  G06V10/44 ,  G06V10/46 ,  G06V10/752 ,  G06V10/772 ,  G06T2207/20081 ,  G06T2207/30164 ,  G06V2201/06

Abstract: A system and method for scoring trained probes for use in analyzing one or more candidate poses of a runtime image is provided. A set of probes with location and gradient direction based on a trained model are applied to one or more candidate poses based upon a runtime image. The applied probes each respectively include a discrete set of position offsets with respect to the gradient direction thereof. A match score is computed for each of the probes, which includes estimating a best match position for each of the probes respectively relative to one of the offsets thereof, and generating a set of individual probe scores for each of the probes, respectively at the estimated best match position.

公开(公告)号：US12033373B2

公开(公告)日：2024-07-09

申请号：US17490709

申请日：2021-09-30

Applicant: HYUNDAI MOTOR COMPANY and KIA CORPORATION

Inventor： David Crawford Gibbon ,  Zhu Liu ,  Eric Zavesky ,  Benjamin J. Stern

IPC: G06V10/772 ,  G06F18/28 ,  G06T5/92 ,  G06V20/10 ,  G06V40/16

CPC classification number: G06V10/772 ,  G06F18/28 ,  G06T5/92 ,  G06V20/10 ,  G06V40/16 ,  G06V40/172 ,  G06K2207/1012 ,  G06T2207/20208

Abstract: Aspects of the subject disclosure may include, for example, receiving first image data for a first image from an electronic device, the first image data including first information associated with a first object identified by the electronic device as being in the first image. Second image data for a second image is received from the electronic device subsequent to receiving the first image data. The second image data includes second information associated with a second object identified by the electronic device as being in the second image. Location information is determined based on the first information and the second information and sent to the electronic device. Other embodiments are disclosed.

公开(公告)号：US12019716B2

公开(公告)日：2024-06-25

申请号：US16661894

申请日：2019-10-23

Applicant: Avago Technologies International Sales Pte. Limited

Inventor： Yong Li ,  Xuemin Chen ,  Brett Tischler ,  Prashant Katre

IPC: G06Q30/00 ,  G06F18/22 ,  G06F18/28 ,  G06F21/12 ,  G06F21/16 ,  G06N5/04 ,  G06N20/00 ,  G06V10/772 ,  G06V20/40

CPC classification number: G06F21/16 ,  G06F18/22 ,  G06F18/28 ,  G06F21/125 ,  G06N5/04 ,  G06N20/00 ,  G06V10/772 ,  G06V20/46

Abstract: A system for multimedia content recognition includes a cloud server and a media client including a silicon-on-chip (SoC) device to communicate with the cloud server via a network. The SoC device includes a local area network (LAN) interface to receive media content from a media source and a media monitor to analyze the received media content and to generate signature information for transmission to the cloud server or for a local analysis. The SoC device further includes an inference engine to locally analyze the signature information to detect an unauthorized access.

公开(公告)号：US20240185574A1

公开(公告)日：2024-06-06

申请号：US18137108

申请日：2023-04-20

Applicant: Simple Intelligence, Inc.

Inventor： Rahul Suresh

IPC: G06V10/772 ,  G06F16/951 ,  G06V10/26 ,  G06V10/762 ,  G06V10/774 ,  G06V10/82 ,  G06V20/70

CPC classification number: G06V10/772 ,  G06F16/951 ,  G06V10/26 ,  G06V10/762 ,  G06V10/774 ,  G06V10/82 ,  G06V20/70

Abstract: The present disclosure relates generally to artificial intelligence (AI), machine learning (ML), and deep learning technologies. More specifically, the disclosure relates to a vehicle image composite system that employs computer vision (CV) along with a Generative Adversarial Network (GAN) to generate realistic composite car images. For example, in one or more embodiments, the composite car image generator system trains a Convolutional Neural Network (CNN) to learn the Make Model Year parameters of all vehicle images provided. Once trained, the determined Make Model Year parameters of the vehicles allow the CNN to produce realistic composite images of a vehicle of any make, model, year, and trim level.