-
公开(公告)号：US20220180622A1
公开(公告)日：2022-06-09
申请号：US17442697
申请日：2020-07-02
Inventor： Zhaoxiang ZHANG , Tieniu TAN , Chunfeng SONG , Junsong FAN
IPC: G06V10/764 , G06T7/174 , G06V10/774 , G06V10/776 , G06V20/70 , G06V10/40
Abstract: A weakly supervised image semantic segmentation method based on an intra-class discriminator includes: constructing two levels of intra-class discriminators for each image-level class to determine whether pixels belonging to the image class belong to a target foreground or a background, and using weakly supervised data for training; generating a pixel-level image class label based on the two levels of intra-class discriminators, and generating and outputting a semantic segmentation result; and further training an image semantic segmentation module or network by using the label to obtain a final semantic segmentation model for an unlabeled input image. By means of the new method, intra-class image information implied in a feature code is fully mined, foreground and background pixels are accurately distinguished, and performance of a weakly supervised semantic segmentation model is significantly improved under the condition of only relying on an image-level annotation.
-
公开(公告)号：US11305817B2
公开(公告)日：2022-04-19
申请号：US16959414
申请日：2018-12-21
Inventor： Jinguo Liu , Chenguang Chang , Qingliang Song
Abstract: The present invention belongs to the field of robots, and relates to a soft ground crawling robot. Front wheels are connected on both sides of the front end of a body shell; a universal wheel is arranged on the rear end; rotary stepping motors are installed on both sides of a supporting assembly; an output shaft of the rotary stepping motor on each side passes through the body shell and then is connected with the front wheel on the same side; a swinging stepping motor is installed on the supporting assembly; the output shaft is connected with a connecting plate; the connecting plate is connected with the body shell through a connecting shaft; the body shell is driven to swing by the swinging stepping motor; lower end covers are rotatably connected on both sides of the body shell; a wheel bracket is connected to the rotary stepping motor on each side; and the wheel bracket on each side is connected with the lower end cover on the same side. The crawling robot of the present invention can crawl on all soft ground while moving in a plane, and overcome the phenomenon that some existing crawling robots cannot walk and work on the soft ground such as sand due to sinking.
-
73.发明授权公开(公告)号：US11300664B1
公开(公告)日：2022-04-12
申请号：US17450693
申请日：2021-10-13
Inventor： Zhiqiang Cao , Zhonghui Li , Jinli Kang , Min Tan , Shuang Liang , Chengpeng Wang
Abstract: A LiDAR odometry method based on the directed geometric point and sparse frame includes: obtaining original three-dimensional point cloud data of a surrounding environment in the coordinate system of the current frame of the LiDAR sensor; constructing and converting the first directed geometric point set and first environmental directed geometric point set into the second directed geometric point set and second environmental directed geometric point set in the world coordinate system, respectively; calculating candidate associated points in the second environmental directed geometric point set in association with each point in the second directed geometric point set, and determining the best associated point of each point in the second directed geometric point set; constructing and solving a pose optimization function to obtain an optimized pose from the coordinate system of the current frame of the LiDAR sensor to the world coordinate system; and updating directed geometric point sets of sparse frames.
-
公开(公告)号：US11243166B2
公开(公告)日：2022-02-08
申请号：US16545023
申请日：2019-08-20
Inventor： Jie Tian , Zhenhua Hu , Caiguang Cao , Zeyu Zhang , Meishan Cai
Abstract: An intraoperative near-infrared-I and near-infrared-II multi-spectral fluorescent navigation system and method of using same includes a light source module for emitting white light and excitation light for illuminating tissue to be tested to generate an emission light. An optical information collection module includes a white light camera for collecting the white light image, and near infrared-I and near infrared-II fluorescence cameras for collecting the near infrared-I and near infrared-II fluorescence images. A central control module is coupled to the light source and the optical information collection modules. An image processing unit pre-processes the white light image, and the near infrared-I and near infrared-II fluorescence images, for de-noising and enhancement. The image processing unit performs a pseudo-color mapping on the images to obtain pseudo-color superimposed images of the near infrared-I and near infrared-II for a surgical region, and completes imaging of the intraoperative near-infrared-I and near-infrared-II multi-spectral fluorescent navigation system.
-
公开(公告)号：US11227161B1
公开(公告)日：2022-01-18
申请号：US17471485
申请日：2021-09-10
Inventor： Jianhua Tao , Yu He , Bin Liu , Licai Sun
Abstract: A physiological signal prediction method includes: collecting a video file, the video file containing long-term videos, and contents of the video file containing data for a face of a single person and true physiological signal data; segmenting a single long-term video into multiple short-term video clips; extracting, by using each frame of image in each of the short-term video clips, features of interested regions for identifying physiological signals so as to form features of interested regions of a single frame; splicing, for each of the short-term video clips, features of interested regions of all fixed frames corresponding to the short-term video clip into features of interested regions of a multi-frame video, and converting the features of the interested regions of the multi-frame video into a spatio-temporal graph; inputting the spatio-temporal graph into a deep learning model for training, and using the trained deep learning model to predict physiological signal parameters.
-
Patent Agency Ranking
公开(公告)号：US11209843B1
公开(公告)日：2021-12-28
申请号：US17372554
申请日：2021-07-12
Inventor： Hongjun Yang , Min Tan , Zengguang Hou , Junzhi Yu , Long Cheng , Zhengxing Wu , Wei He , Zhijie Liu , Tairen Sun
Abstract: A control method, system and device for a flexible carbon cantilever beam actuated by a smart material is provided. The new control method, system and device aims to solve the problems of control overflow and instability that are likely to occur in the distributed parameter system constructed in the prior art. The method includes: acquiring an elastic displacement of the flexible carbon cantilever beam in real time as input data; and obtaining, based on the input data, a control torque through a distributed parameter model constructed in advance, and performing vibration control on the flexible carbon cantilever beam. The new control method, system and device improves the control accuracy and stability of the distributed parameter system.
-
公开(公告)号：US11208186B2
公开(公告)日：2021-12-28
申请号：US17279099
申请日：2020-04-16
Inventor： Junzhi Yu , Zhengxing Wu , Di Chen , Min Tan
Abstract: A water-air amphibious cross-medium bio-robotic flying fish includes a body, pitching pectoral fins, variable-structure pectoral fins, a caudal propulsion module, a sensor module and a controller. The caudal propulsion module is controlled to achieve underwater fish-like body-caudal fin (BCF) propulsion, and the variable-structure pectoral fins is adjusted to achieve air gliding and fast splash-down diving motions of the bio-robotic flying fish. The coordination between the caudal propulsion module and the pitching pectoral fins is controlled to achieve the motion of leaping out of water during water-air cross-medium transition. The ambient environment is detected by the sensor module, and the motion mode of the bio-robotic flying fish is controlled by the controller.
-
78.发明授权公开(公告)号：US11175490B1
公开(公告)日：2021-11-16
申请号：US17322926
申请日：2021-05-18
Inventor： Guibin Bian , Jinlian Song , Wenhao He , Zhen Li , Haitao Song
IPC: H04N7/18 , G02B21/36 , G02B30/24 , H04N13/239 , H04N13/296 , H04N13/341 , H04N13/128 , H04N13/167 , H04N13/00
Abstract: A shutter-type adaptive three-dimensional (3D) display system based on medical microscopic imaging is provided. In the system, a binocular camera is connected to an operating microscope optical system through a beam splitter, and is configured to acquire binocular image data. A control module is configured to input control instructions, where the control instructions include a first type of control instructions and a second type of control instructions. An image processing module calculates image parallaxes based on the binocular image data under the first type of control instructions, and transmits the image parallaxes to the binocular camera to adjust offsets. A display module is configured to generate display data based on the binocular image data and display the display data through a display device for viewing with 3D glasses. The operating microscope optical system is configured to perform an adjustment based on the control instructions.
-
公开(公告)号：US20210268231A1
公开(公告)日：2021-09-02
申请号：US17044513
申请日：2019-11-12
Inventor： Xiaoliang XIE , Zengguang HOU , Lingwu MENG , Zhenqiu FENG , Xiaohu ZHOU , Shiqi LIU
Abstract: A vascular interventional instrument control device capable of operating double guide wires or balloons, includes a main finger assembly, a first sub-finger assembly, a second sub-finger assembly, a driving assembly and a clamping assembly, wherein the main finger assembly, the sub-finger assembly, the driving assembly and the clamping assembly are separately installed to the body structural member and separately connected to the controller through a communication link, one of the operated double guide wires or balloons is clamped by the main finger assembly and the first sub-finger assembly, and the other is clamped by the main finger assembly and the second sub-finger assembly; and the operated double guide wires or double balloons can be separately moved axially or rotated about its own axial direction.
-
公开(公告)号：US20210247737A1
公开(公告)日：2021-08-12
申请号：US17258748
申请日：2019-09-16
Applicant: INSTITUTE OF AUTOMATION, CHINESE ACADEMY OF SCIENCES , BEIJING TEN DIMENSIONS TECHNOLOGY CO., LTD.
Inventor： Zhen SHEN , Gang XIONG , Yuqing LI , Hang GAO , Yi XIE , Meihua ZHAO , Chao GUO , Xiuqin SHANG , Xisong DONG , Zhengpeng WU , Li WAN , Feiyue WANG
IPC: G05B19/4099 , G06N3/08 , G05B19/404
Abstract: A neural network-based error compensation method for 3D printing includes: compensating an input model by a deformation network/inverse deformation network constructed and trained according to a 3D printing deformation function/inverse deformation function, and performing the 3D printing based on the compensated model. Training samples of the deformation network/inverse deformation network include to-be-printed model samples and printed model samples. The deformation network constructed according to the 3D printing deformation function is marked as a first network. During training of the first network, the to-be-printed model samples are used as real input models, and the printed model samples are used as real output models. The inverse deformation network constructed according to the 3D printing inverse deformation function is marked as a second network. During training of the second network, the printed model samples are used as real input models, and the to-be-printed model samples are used as real output models.
-
-
-
-
-
-
-
-
-
Search Results
243
records
(took 0.024 seconds)
