公开(公告)号：US11868892B2

公开(公告)日：2024-01-09

申请号：US17887359

申请日：2022-08-12

申请人： INTEL CORPORATION

发明人： Furkan Isikdogan ,  Bhavin V. Nayak ,  Joao Peralta Moreira ,  Chyuan-Tyng Wu ,  Gilad Michael

IPC分类号： G06N3/08 ,  G06V10/70 ,  G06V10/764 ,  G06N3/063 ,  G06F18/214 ,  G06N3/048 ,  G06V10/774 ,  G06V10/776 ,  G06V10/82

CPC分类号： G06N3/08 ,  G06F18/214 ,  G06N3/048 ,  G06N3/063 ,  G06V10/70 ,  G06V10/764 ,  G06V10/774 ,  G06V10/776 ,  G06V10/82

摘要： An apparatus to facilitate partially-frozen neural networks for efficient computer vision systems is disclosed. The apparatus includes a frozen core to store fixed weights of a machine learning model, one or more trainable cores coupled to the frozen core, the one or more trainable cores comprising multipliers for trainable weights of the machine learning model, and wherein the alpha blending layer includes a trainable alpha blending parameter, and wherein the trainable alpha blending parameter is a function of a trainable parameter, a sigmoid function, and outputs of frozen and trainable blocks in a preceding layer of the machine learning model.

公开(公告)号：US20240005636A1

公开(公告)日：2024-01-04

申请号：US18311728

申请日：2023-05-03

申请人： Bayerische Motoren Werke Aktiengesellschaft

发明人： Etienne Iliffe-Moon

IPC分类号： G06V10/70 ,  G06V10/764 ,  G06V20/59

CPC分类号： G06V10/768 ,  G06V10/764 ,  G06V20/597 ,  G06V2201/02

摘要： A method for a vehicle for estimating a user context includes acquiring, by a camera of the vehicle, image information relating to an output device of a user terminal device. The method also includes obtaining, by performing classification based on the image information, device status information relating to the user terminal device, and estimating the user context based on the device status information.

公开(公告)号：US11860928B2

公开(公告)日：2024-01-02

申请号：US17406893

申请日：2021-08-19

申请人： International Business Machines Corporation

发明人： Xiaoxiao Guo ,  Hui Wu ,  Rogerio Feris

IPC分类号： G06F16/00 ,  G06F16/53 ,  G06F16/332 ,  G06F16/383 ,  G06N3/049 ,  G06F16/535 ,  G06F18/21 ,  G06V10/70 ,  G06V10/82 ,  G06V40/10

CPC分类号： G06F16/53 ,  G06F16/3328 ,  G06F16/3329 ,  G06F16/383 ,  G06F16/535 ,  G06F18/2185 ,  G06N3/049 ,  G06V10/768 ,  G06V10/82 ,  G06V40/103

摘要： A method includes receiving input from a client at least partially specifying one or more characteristics, wherein the initial input includes a seed image and a natural language statement describing a desired change to the seed image; predicting one or more attributes of the seed image by operation of a neural network on the seed image; and parsing the natural language statement to identify desired changes to the one or more attributes of the seed image. The method also includes generating an interim target image by changing the one or more attributes of the seed image, according to the parsed natural language statement; selecting a set of images from an image database for output to the client, each of said set of images being determined to at least partially satisfy the one or more changed attributes of the seed image; and displaying the set of images to the client.

公开(公告)号：US20230409311A1

公开(公告)日：2023-12-21

申请号：US18365347

申请日：2023-08-04

申请人： Microsoft Technology Licensing, LLC

发明人： Yutong LIAO ,  Cheng WU ,  Nicolas Justin LAVIGNE ,  Frederick Douglass CAMPBELL ,  Chan CHAIYOCHLARB ,  Raymond Duane PARSONS ,  Alexander OOT ,  Paul Luo LI ,  Minsuk KANG ,  Abhinav MISHRA

IPC分类号： G06F8/65 ,  G06N3/08 ,  G06F9/54 ,  G06V10/70 ,  G06F18/21 ,  G06N20/00

CPC分类号： G06F8/65 ,  G06N3/08 ,  G06N20/00 ,  G06V10/768 ,  G06F18/217 ,  G06F9/542

摘要： The present disclosure is directed to automated generation and management of update estimates relative to application of an update to a computing device. One or more updated to be applied to a computing device are identified. A trained artificial intelligence (AI) model is applied that is adapted to generate an update estimate predicting an amount of time that is required to apply an update to the computing device. An update estimate is generated based on a contextual analysis that evaluates one or more of: parameters associated with the update; device characteristics of the computing device to be updated; a state of current user activity on the computing device; historical predictions relating to prior update estimates for one or more computing devices (e.g., that comprise the computing device); or a combination thereof. A notification of the update estimate is then automatically generated and caused to be rendered.

公开(公告)号：US20230406297A1

公开(公告)日：2023-12-21

申请号：US18241576

申请日：2023-09-01

申请人： TuSimple, Inc.

发明人： Hongkai YU ,  Zhipeng YAN ,  Panqu WANG ,  Pengfei CHEN

IPC分类号： B60W30/095 ,  G06V20/56 ,  G06V10/44 ,  G06V20/58 ,  G06F18/24 ,  G06F18/214 ,  G06F18/2413 ,  G06V10/25 ,  G06V10/774 ,  G06V10/70

CPC分类号： B60W30/0956 ,  G06V20/56 ,  G06V10/44 ,  G06V20/58 ,  G06F18/24 ,  G06F18/2155 ,  G06F18/2413 ,  G06V10/25 ,  G06V10/7753 ,  G06V10/87 ,  G05D1/0221

摘要： A system and method for vehicle occlusion detection is disclosed. A particular embodiment includes: receiving training image data from a training image data collection system; obtaining ground truth data corresponding to the training image data; performing a training phase to train a plurality of classifiers, a first classifier being trained for processing static images of the training image data, a second classifier being trained for processing image sequences of the training image data; receiving image data from an image data collection system associated with an autonomous vehicle; and performing an operational phase including performing feature extraction on the image data, determining a presence of an extracted feature instance in multiple image frames of the image data by tracing the extracted feature instance back to a previous plurality of N frames relative to a current frame, applying the first trained classifier to the extracted feature instance if the extracted feature instance cannot be determined to be present in multiple image frames of the image data, and applying the second trained classifier to the extracted feature instance if the extracted feature instance can be determined to be present in multiple image frames of the image data.

公开(公告)号：US20230401870A1

公开(公告)日：2023-12-14

申请号：US18455489

申请日：2023-08-24

申请人： DENSO CORPORATION

发明人： Hideaki MISAWA

IPC分类号： G06V20/54 ,  G06V10/70 ,  B60W50/14 ,  B60W60/00

CPC分类号： G06V20/54 ,  G06V10/70 ,  B60W50/14 ,  B60W60/001 ,  B60W2050/146 ,  B60W2540/215 ,  B60W2556/10 ,  B60W2556/20 ,  B60W2556/45

摘要： An autonomous driving system includes a recognition unit configured to recognize an obstacle based on image data obtained by imaging a predetermined region including a road on which an autonomous vehicle travels, the imaging being performed by an imaging device installed at a specific location in the external environment of the autonomous vehicle, a feature quantity distribution creation unit configured to create a feature quantity distribution expressing a distribution of features related to obstacles recognized in the past, and an erroneous recognition judgement unit configured to compare the created feature quantity distribution with a feature quantity of an obstacle recognized as an evaluation subject, to thereby judge whether the obstacle recognized as the evaluation subject is erroneously recognized.

公开(公告)号：US20230394611A1

公开(公告)日：2023-12-07

申请号：US18448147

申请日：2023-08-10

申请人： CHENGDU QINCHUAN IOT TECHNOLOGY CO., LTD.

发明人： Zehua SHAO ,  Haitang XIANG ,  Yong LI ,  Yaqiang QUAN ,  Bin LIU

IPC分类号： G06Q50/26 ,  G06V10/70 ,  G06Q10/04 ,  H04L67/12

CPC分类号： G06Q50/26 ,  G06V10/70 ,  G06Q10/04 ,  H04L67/12

摘要： The present disclosure provides a method and a system for area management in a smart city based on an Internet of Things. The method includes obtaining environmental monitoring data in a target area through a sensor network platform, the environmental monitoring data including at least one of air quality data, weather data, and satellite image data, predicting an air pollution situation in the target area through a regional prediction model based on the environmental monitoring data, the regional prediction model being a machine learning model, and sending prompt information, which is determined based on the air pollution situation in the target area, to a user platform through a service platform, wherein the target area is a hexagonal area; and the regional prediction model includes seven regional models and an air quality determination model; wherein a prediction mode of the regional prediction model is an iterative prediction.

公开(公告)号：US20230394311A1

公开(公告)日：2023-12-07

申请号：US18236864

申请日：2023-08-22

申请人： Advanced Micro Devices, Inc.

发明人： Sudhanva Gurumurthi ,  Abhinav Vishnu

IPC分类号： G06N3/082 ,  G06N20/20 ,  G06T7/90 ,  G06N3/045 ,  G06V10/56 ,  G06V10/82 ,  G06V10/70

CPC分类号： G06N3/082 ,  G06N20/20 ,  G06T7/90 ,  G06N3/045 ,  G06V10/56 ,  G06V10/82 ,  G06V10/87

摘要： A system comprising an electronic device that includes a processor is described. During operation, the processor acquires a full version of a neural network, the neural network including internal elements for processing instances of input image data having a set of color channels. The processor then generates, from the neural network, a set of sub-networks, each sub-network being a separate copy of the neural network with the internal elements for processing at least one of the color channels in instances of input image data removed, so that each sub-network is configured for processing a different set of one or more color channels in instances of input image data. The processor next provides the sub-networks for processing instances of input image data—and may itself use the sub-networks for processing instances of input image data.

公开(公告)号：US20230388626A1

公开(公告)日：2023-11-30

申请号：US18323167

申请日：2023-05-24

申请人： CARIAD SE

发明人： Octavian Tirla

IPC分类号： H04N23/60 ,  G06V20/59 ,  G06V10/12 ,  G06V10/70 ,  G06V10/80

CPC分类号： H04N23/64 ,  G06V20/59 ,  G06V10/12 ,  G06V10/87 ,  G06V10/803

摘要： The disclosure relates to a method for operating a vehicle interior monitoring system including at least one camera unit. A control device sets an adjustable camera parameter of the camera unit by a camera-specific control command in the camera unit and receives and evaluates at least one image recorded by the camera unit, and a result of the evaluation is output as a camera-specific result datum. The control device generates a general control command that sets the adjustable camera parameter of the camera unit, the general control command is converted based on a camera configuration of the camera unit saved in the control device into the camera-specific control command, the camera-specific result datum is converted based on the camera configuration of the camera unit saved in the control device into a general result datum, and the general result datum of the camera unit is provided to a data fusion device.

公开(公告)号：US20230388445A1

公开(公告)日：2023-11-30

申请号：US17826639

申请日：2022-05-27

申请人： MOTOROLA MOBILITY LLC

发明人： AMIT KUMAR AGRAWAL ,  JEFFREY S VANHOOF ,  BILL RYAN ,  DANIEL M VACURA

IPC分类号： H04N5/265 ,  G06V20/40 ,  G06V10/25 ,  G06V40/16 ,  G06T11/00 ,  H04N5/262 ,  G06V10/10 ,  G06T7/12 ,  G06V20/62 ,  G06V10/70

CPC分类号： H04N5/265 ,  G06V20/49 ,  G06V20/41 ,  G06V10/25 ,  G06V40/161 ,  G06T11/00 ,  H04N5/2628 ,  G06V10/16 ,  G06T7/12 ,  G06V20/62 ,  G06V10/70 ,  G06T2207/20021 ,  G06T2207/20132 ,  G06T2207/30201 ,  G06T2207/10016 ,  G06T2200/24 ,  G06T2207/20212

摘要： A method provides non-mirrored preview of text-based demonstration objects in a locally captured video stream. The method includes identifying a demonstration object within the video stream. The demonstration object contains content that is best when pre-viewed in a non-mirrored orientation. The method includes spatially segmenting the video image into two or more segments, including a demonstration object preview (DOP) segment that encompasses the defined area with the demonstration object and at least one second segment encompassing a remaining portion of the video image. The method includes presenting a preview of the video image on the display device with the remaining portion of the video image mirrored within the preview and at least the DOP segment presented without mirroring in a correct spatial location relative to the remaining portion of the video image. A person presenting the demonstration object receives a non-mirrored presentation preview of the demonstration object.