公开(公告)号：US20210209339A1

公开(公告)日：2021-07-08

申请号：US17059967

申请日：2018-08-31

Applicant: Intel Corporation

Inventor： Ganmei YOU ,  Zhigang WANG ,  Dawei WANG

IPC: G06K9/00 ,  G06K9/62 ,  G06K9/46 ,  G06N3/04 ,  G06N3/08

Abstract: Techniques related to training and implementing convolutional neural networks for object recognition are discussed. Such techniques may include applying, at a first convolutional layer of the convolutional neural network, 3D filters of different spatial sizes to an 3D input image segment to generate multi-scale feature maps such that each feature map has a pathway to fully connected layers of the convolutional neural network, which generate object recognition data corresponding to the 3D input image segment.

公开(公告)号：US20240029300A1

公开(公告)日：2024-01-25

申请号：US18254181

申请日：2020-12-25

Applicant: INTEL CORPORATION

Inventor： Xuesong SHI ,  Yuxin TIAN ,  Sangeeta GHANGAM ,  Dawei WANG

IPC: G06T7/73 ,  G06T7/60 ,  G06T3/00 ,  G06T1/00

CPC classification number: G06T7/74 ,  G06T7/60 ,  G06T3/0075 ,  G06T1/0014 ,  G06T2207/10024 ,  G06T2207/10028

Abstract: A method for re-localization of the robot may include retrieving, for each of keyframes in a keyframe database of the robot, image features and a pose of the keyframe, the image features of the keyframe comprising a global descriptor and local descriptors of the keyframe (210); extracting image features of a current frame captured by the robot, the image features of the current frame comprising a global descriptor and local descriptors of the current frame (220); determining one or more rough matching frames from the keyframes based on comparison between the global descriptor of each keyframe and the global descriptor of the current frame (230); determining a final matching frame from the one or more rough matching frames based on comparison between the local descriptors of each rough matching frame and the local descriptors of the current frame (240); and calculating a pose of the current frame based on a pose of the current frame based on a pose of the final matching frame (250).

公开(公告)号：US20210302168A1

公开(公告)日：2021-09-30

申请号：US16831104

申请日：2020-03-26

Applicant: Intel Corporation

Inventor： Xiaolong LIU ,  Zhigang WANG ,  Dawei WANG ,  Haitao JI ,  Qianying ZHU ,  Fei LI ,  Ignacio J ALVAREZ

IPC: G01C21/30 ,  G06K9/00 ,  G01C21/36

Abstract: The present disclosure relates to real-time localization error correction of an autonomous vehicle (AV). A processor for real-time localization error correction of the AV is provided. The processor is configured to retrieve a reference landmark around the AV from a map aggregating server (MAS), wherein the AV is configured to interact with the MAS for real-time localization; detect, in real time, a ground truth landmark corresponding to the reference landmark, according to image data captured by one or more image capture devices installed on the AV; and determine a deviation between the ground truth landmark and the reference landmark as a real-time correction value for the real-time localization of the AV.

公开(公告)号：US20240112035A1

公开(公告)日：2024-04-04

申请号：US18527490

申请日：2023-12-04

Applicant: Intel Corporation

Inventor： Ganmei YOU ,  Zhigang WANG ,  Dawei WANG

IPC: G06N3/084 ,  G06F18/213 ,  G06N3/04 ,  G06V10/44 ,  G06V10/764 ,  G06V10/82 ,  G06V20/56 ,  G06V20/58 ,  G06V20/64

CPC classification number: G06N3/084 ,  G06F18/213 ,  G06N3/04 ,  G06V10/454 ,  G06V10/764 ,  G06V10/82 ,  G06V20/56 ,  G06V20/58 ,  G06V20/64

Abstract: Techniques related to training and implementing convolutional neural networks for object recognition are discussed. Such techniques may include applying, at a first convolutional layer of the convolutional neural network, 3D filters of different spatial sizes to an 3D input image segment to generate multi-scale feature maps such that each feature map has a pathway to fully connected layers of the convolutional neural network, which generate object recognition data corresponding to the 3D input image segment.