公开(公告)号：US20240364840A1

公开(公告)日：2024-10-31

申请号：US18628311

申请日：2024-04-05

Applicant: Zoom Video Communications, Inc.

Inventor： Lin Han ,  Nitasha Walia

IPC: H04N5/272 ,  G06T7/12 ,  G06T7/194 ,  G06T7/20 ,  H04N7/15

CPC classification number: H04N5/272 ,  G06T7/12 ,  G06T7/194 ,  G06T7/20 ,  H04N7/15 ,  G06T2207/10016 ,  G06T2207/30196

Abstract: One example system for displaying immersive scenes includes a processor and at least one memory device. The memory device includes instructions that are executable by the processor to cause the processor to receive a collection of metadata associated with an immersive scene, identify each of a plurality of properties of the immersive scene based on the collection of metadata, receive a dynamic immersive background, receive a plurality of video streams associated with a video conference, and display each of the plurality of video streams in the immersive scene based at least in part of the plurality of properties of the immersive scene and on the dynamic immersive background.

公开(公告)号：US20240362791A1

公开(公告)日：2024-10-31

申请号：US18307353

申请日：2023-04-26

Applicant: Adobe Inc.

Inventor： Yuqian Zhou ,  Chuong Huynh ,  Connelly Barnes ,  Elya Shechtman ,  Sohrab Amirghodsi ,  Zhe Lin

IPC: G06T7/12 ,  G06F3/04883 ,  G06V10/44 ,  G06V10/74 ,  G06V10/80

CPC classification number: G06T7/12 ,  G06F3/04883 ,  G06V10/44 ,  G06V10/761 ,  G06V10/806 ,  G06T2207/20101 ,  G06V2201/07

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods for utilizing machine learning to generate a mask for an object portrayed in a digital image. For example, in some embodiments, the disclosed systems utilize a neural network to generate an image feature representation from the digital image. The disclosed systems can receive a selection input identifying one or more pixels corresponding to the object. In addition, in some implementations, the disclosed systems generate a modified feature representation by integrating the selection input into the image feature representation. Moreover, in one or more embodiments, the disclosed systems utilize an additional neural network to generate a plurality of masking proposals for the object from the modified feature representation. Furthermore, in some embodiments, the disclosed systems utilize a further neural network to generate the mask for the object from the modified feature representation and/or the masking proposals.

公开(公告)号：US20240354952A1

公开(公告)日：2024-10-24

申请号：US18759821

申请日：2024-06-29

Applicant: SHANGHAI UNITED IMAGING INTELLIGENCE CO., LTD. ,  UNITED IMAGING INTELLIGENCE (BEIJING) CO., LTD.

Inventor： Zirong CHEN ,  Dijia WU ,  Pei DONG

IPC: G06T7/00 ,  G06T7/12 ,  G06V10/44 ,  G06V20/70 ,  G16H30/40

CPC classification number: G06T7/0012 ,  G06T7/12 ,  G06V10/44 ,  G06V20/70 ,  G16H30/40 ,  G06T2207/30048 ,  G06T2207/30101

Abstract: A system and a method for bypass vessel reconstruction may be provided. A target image including at least a cardiac region of a subject may be obtained. A first segmentation result and a second segmentation result may be determined based on the target image. The first segmentation result and the second segmentation result may respectively indicate the heart and vessels of the subject segmented from the target image. A target segment result may be determined using a vessel segment model based on the first segmentation result and the second segmentation result. The target segment result may include a plurality of segment labels of a plurality of points on the vessels of the subject, and the plurality of segment labels may include a segment label corresponding to bypass vessels. Further, data relating to one or more bypass vessels of the subject may be determined based on the target segment result.

公开(公告)号：US20240354945A1

公开(公告)日：2024-10-24

申请号：US18640357

申请日：2024-04-19

Applicant: MEDICALIP CO., LTD.

Inventor： Sang Joon Park ,  Jong Min Kim ,  Joseph Nathanael Witanto

IPC: G06T7/00 ,  G06T7/12 ,  G06V10/60 ,  G16H30/40

CPC classification number: G06T7/0012 ,  G06T7/12 ,  G06V10/60 ,  G16H30/40 ,  G06T2207/10081 ,  G06T2207/10088 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/30061

Abstract: Provided are a medical image translation method and apparatus. The medical image translation apparatus receives a first medical image and translates the first medical image into a second medical image through an image translation model. The first medical image is a two-dimensional (2D) medical image, and the second medical image is a 2D medical image obtained by reconstructing a brightness value of each of pixels of the first medical image while maintaining a structure shown in the first medical image. The image translation model is a model implemented with an artificial neural network that reflects and outputs a feature of a brightness value of each of pixels of a reference image trained in a training process in a 2D medical image.

公开(公告)号：US20240331165A1

公开(公告)日：2024-10-03

申请号：US18579793

申请日：2022-04-28

Applicant: ZHEJIANG UNIVERSITY

Inventor： JIANWEI YIN ,  YUXIANG CAI ,  YINGCHUN YANG ,  YONGHENG SHANG ,  ZHENQIAN CHEN ,  ZHENGWEI SHEN

IPC: G06T7/12 ,  G06T7/11

CPC classification number: G06T7/12 ,  G06T7/11 ,  G06T2207/20081

Abstract: A cross-domain remote sensing image semantic segmentation method based on iterative intra-domain adaptation and self-training, includes training source-target inter-domain domain adaptation models, generating target domain category segmentation probabilities and pseudo labels, sorting target domain image segmentation probability credibility scores, training target intra-domain iterative domain adaptation models, and generating target domain segmentation results. The invention utilizes source-target domain inter-domain adaptation to reduce source-target domain inter-domain differences, and, utilizes target intra-domain adaptation to reduce target intra-domain differences and improve the accuracy of cross-domain remote sensing image semantic segmentation models; furthermore, by proposed classifying and sorting target domain images based on segmentation probability credibility, prediction results with good segmentation effects are selected as pseudo labels; meanwhile, a new pseudo label screening strategy is proposed to remove pixel points that are highly likely to be erroneous from pseudo labels, thereby avoiding the impact caused by erroneous pseudo labels during self-training within target domains.

公开(公告)号：US20240331164A1

公开(公告)日：2024-10-03

申请号：US18194164

申请日：2023-03-31

Applicant: KING FAHD UNIVERSITY OF PETROLEUM AND MINERALS

Inventor： Ziyad Adel ALSWAIDAN ,  Khalil Hedi CHIKHAOUI ,  Mohammad Hashem M. Kenana SHULLAR ,  Motaz ALFARRAJ

IPC: G06T7/12 ,  G06T3/00 ,  G06T7/11 ,  G06V10/46

CPC classification number: G06T7/12 ,  G06T3/18 ,  G06T7/11 ,  G06V10/462 ,  G06T2207/20084 ,  G06T2207/20132 ,  G06V2201/07

Abstract: A method of image retargeting including segmenting an input image into a plurality of cells and detecting elements in the input image by a deep learning-based semantic segmentation network. The method includes generating a segmentation map according to a first function, generating a saliency map according to a second function and generating a variance map according to a third function. The method includes constructing an importance map according to a fourth function. The importance map is based on the segmentation map, the saliency map, and the variance map. The method includes inputting the importance map into a distortion threshold function. The distortion threshold function calculates a distortion threshold. The method includes calculating a distortion measure and determining a modification to be applied to each of the plurality of cells. The method includes applying the modification to each of the plurality of cells and outputting a retargeted image.

公开(公告)号：US20240312023A1

公开(公告)日：2024-09-19

申请号：US18211244

申请日：2023-06-17

Applicant: Inventec (Pudong) Technology Corp. ,  Inventec Corporation

Inventor： Ping-Jung Lu ,  I-Chang Wang

IPC: G06T7/12 ,  G06F30/10

CPC classification number: G06T7/12 ,  G06F30/10 ,  G06T2207/20021 ,  G06T2207/20036

Abstract: A contour line processing method is provided, which includes obtaining a computer-aided design image file and detecting a plurality of line segments from the computer-aided design image file, performing a first morphological operation on the plurality of line segments to generate a plurality of blocks, performing a union operation on the plurality of blocks to generate a union block, performing a second morphological operation on the union block to generate a processed block, and determining a target contour block according to the processed block.

公开(公告)号：US20240312013A1

公开(公告)日：2024-09-19

申请号：US18676487

申请日：2024-05-28

Applicant: SHANGHAI UNITED IMAGING HEALTHCARE CO., LTD.

Inventor： Liang XU ,  Jian ZHONG ,  Juan FENG ,  Yan'ge MA ,  Zhou YUAN

IPC: G06T7/00 ,  A61B6/00 ,  G06T7/12 ,  G06T7/73 ,  G06V10/77 ,  G06V10/774 ,  G06V10/82 ,  G06V40/10 ,  G16H30/40

CPC classification number: G06T7/0012 ,  A61B6/488 ,  G06T7/12 ,  G06T7/73 ,  G06V10/7715 ,  G06V10/774 ,  G06V10/82 ,  G06V40/10 ,  G16H30/40 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/30004 ,  G06T2207/30196 ,  G06V2201/03

Abstract: A method and a system for medical imaging may be provided. A first optical image of a target subject that includes a target region to be scanned or treated by a medical device may be obtained. At least one body part boundary and at least one feature point of the target subject may be identified using at least one target recognition model from the first optical image. An image region corresponding to the target region of the target subject may be identified from the first optical image based on the at least one body part boundary and the at least one feature point. At least one first edge of the image region may be determined based on the at least one body part boundary, and at least one second edge of the image region may be determined based on the at least one feature point.

公开(公告)号：US12094310B2

公开(公告)日：2024-09-17

申请号：US17604190

申请日：2019-04-17

Applicant: HENDRICKS CORP PTE LTD

Inventor： Michael Michael ,  Chuan Hun Ye

IPC: G06K9/00 ,  G06T7/12 ,  G06T7/269 ,  G06V10/44 ,  G06V20/00 ,  G06V20/50 ,  G08B17/10

CPC classification number: G08B17/10 ,  G06T7/12 ,  G06T7/269 ,  G06V10/44 ,  G06V20/00 ,  G06V20/50 ,  G06T2207/10016 ,  G06T2207/10024

Abstract: This invention relates to a method and system for detecting fire and smoke. The system comprises a processor, a memory and instructions stored on the memory and executable by the processor to: receive a sequence of images from a plurality of cameras; sampling the sequence of images at a certain interval; process each of the sampled images to form a first processed image and a second processed image; extract Binary Large Objects (BLOBs) from first processed image and contours of objects from the second processed image; analyse the BLOB for smoke and the contours of objects for fire; and determine smoke in response to analysing smoke from the BLOBs and fire in response to analysing fire from the contours of objects.

公开(公告)号：US12094076B2

公开(公告)日：2024-09-17

申请号：US17451790

申请日：2021-10-21

Applicant: Samsung Electronics Co., Ltd.

Inventor： Junkeun Yoon ,  Kideok Kim ,  Minsoo Kim ,  Jeonghyun Kim

IPC: G06T3/40 ,  G06T7/12

CPC classification number: G06T3/40 ,  G06T7/12 ,  G06T2207/10024 ,  G06T2207/20021 ,  G06T2207/20221

Abstract: An electronic device and a controlling method of the electronic device is disclosed. Specifically, the electronic device according to the disclosure may identify, based on receiving a user input for expanding an outer area of a first image, a plurality of objects included in the first image, and obtain a first segmentation image including segmentation information on areas corresponding to the respective objects, obtain a second segmentation image in which an outer area of the first segmentation image is expanded based on the segmentation information, obtain a second image in which segmentation information included in the second segmentation image is converted to RGB information, obtain a third image by reflecting a feature of the first image to the second image based on the segmentation information, and provide the obtained third image.