Patents by Inventor Yehui YANG
Yehui YANG has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 11735315Abstract: Embodiments of the present disclosure disclose a method, apparatus, and device for fusing features applied to small target detection, and a storage medium, relate to the field of computer vision technology. A particular embodiment of the method for fusing features applied to small target detection comprises: acquiring feature maps output by convolutional layers in a Backbone network; performing convolution on the feature maps to obtain input feature maps of feature layers, the feature layers representing resolutions of the input feature maps; and fusing, based on densely connection feature pyramid network features, the input feature maps of each feature layer to obtain output feature maps of the feature layer. Since no additional convolutional layer is introduced for feature fusion, the detection performance for small targets may be enhanced without additional parameters, and the detection ability for small targets may be improved with computing resource constraints.Type: GrantFiled: March 26, 2021Date of Patent: August 22, 2023Assignee: Beijing Baidu Netcom Science and Technology Co., Ltd.Inventors: Binghong Wu, Yehui Yang, Yanwu Xu, Lei Wang
-
Publication number: 20230195839Abstract: Technical solutions relate to the field of artificial intelligence such as deep learning, computer vision and intelligent imaging. A method may includes during training of a one-stage object detecting model, acquiring values of a loss function corresponding to feature maps at different scales respectively in the case that classification loss calculation is required, and the loss function is a focal loss function; and determining a final value of the loss function according to the acquired values of the loss function, and training the one-stage object detecting model according to the final value of the loss function.Type: ApplicationFiled: December 20, 2021Publication date: June 22, 2023Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.Inventors: Binghong WU, Yehui YANG, DaLU YANG, Yanwu XU, Lei WANG, Qian LI
-
Patent number: 11436447Abstract: A target detection method a is provided, which relates to the fields of deep learning, computer vision, and artificial intelligence. The method comprises: classifying, by using a first classification model, a plurality of image patches comprised in an input image, to obtain one or more candidate image patches, in the plurality of image patches, that are preliminarily classified as comprising a target; extracting a corresponding salience area for each candidate image patch; constructing a corresponding target feature vector for each candidate image patch based on the corresponding salience area for each candidate image patch; and classifying, by using a second classification model, the target feature vector to determine whether each candidate image patch comprises the target.Type: GrantFiled: September 30, 2020Date of Patent: September 6, 2022Assignee: BEIJING BAIDU NETCOM SCIENCE AND TECHNOLOGY CO., LTD.Inventors: Yehui Yang, Lei Wang, Yanwu Xu
-
Patent number: 11416307Abstract: A device and method for automatically allocating computing resources is disclosed herein.Type: GrantFiled: March 18, 2020Date of Patent: August 16, 2022Assignee: ALIBABA GROUP HOLDING LIMITEDInventors: Lanlan Cong, Heshan Lin, Yehui Yang
-
Patent number: 11379980Abstract: The present application discloses an image processing method, an apparatus, an electronic device and a storage medium. A specific implementation is: acquiring an image to be processed; acquiring a grading array according to the image to be processed and a grading network model, where the grading network model is a model pre-trained according to mixed samples, the number of elements contained in the grading array is C?1, C is the number of lesion grades, C lesion grades include one lesion grade without lesion and C?1 lesion grades with lesion, and a kth element in the grading array is a probability of a lesion grade corresponding to the image to be processed being greater than or equal to a kth lesion grade, where 1?k?C?1, and k is an integer; determining the lesion grade corresponding to the image to be processed according to the grading array.Type: GrantFiled: November 13, 2020Date of Patent: July 5, 2022Inventors: Fangxin Shang, Yehui Yang, Lei Wang, Yanwu Xu
-
Patent number: 11232560Abstract: Embodiments of the present disclosure provide a method and apparatus for processing a fundus image. The method may include: acquiring a target fundus image; dividing the target fundus image into at least two first image blocks; inputting a first image block into a pre-trained deep learning model, to obtain a first output value; and determining, based on the first output value and a threshold, whether the first image block is the fundus image block containing a predetermined type of image region.Type: GrantFiled: December 2, 2019Date of Patent: January 25, 2022Inventors: Yehui Yang, Yanwu Xu, Lei Wang, Yan Huang
-
Publication number: 20210406586Abstract: An image classification method and apparatus, and a style transfer model training method and apparatus are provided, which are relate to the field of deep learning, cloud computing and computer vision in artificial intelligence. The image classification method comprises: inputting an image of a first style into a style transfer model, to obtain an image of a second style corresponding to the image of the first style; and inputting the image of the second style into an image classification model, to obtain a classification result of the image of the second style, wherein the style transfer model is obtained through training on the basis of a sample image of the first style and a sample image of the second style; and the image classification model is obtained through training on the basis of the sample image of the second style.Type: ApplicationFiled: December 31, 2020Publication date: December 30, 2021Inventors: Dalu YANG, Yehui YANG, Lei WANG, Yanwu XU
-
Publication number: 20210406616Abstract: A target detection method a is provided, which relates to the fields of deep learning, computer vision, and artificial intelligence. The method comprises: classifying, by using a first classification model, a plurality of image patches comprised in an input image, to obtain one or more candidate image patches, in the plurality of image patches, that are preliminarily classified as comprising a target; extracting a corresponding salience area for each candidate image patch; constructing a corresponding target feature vector for each candidate image patch based on the corresponding salience area for each candidate image patch; and classifying, by using a second classification model, the target feature vector to determine whether each candidate image patch comprises the target.Type: ApplicationFiled: September 30, 2020Publication date: December 30, 2021Inventors: Yehui YANG, Lei WANG, Yanwu XU
-
Publication number: 20210224581Abstract: Embodiments of the present disclosure disclose a method, apparatus, and device for fusing features applied to small target detection, and a storage medium, relate to the field of computer vision technology. A particular embodiment of the method for fusing features applied to small target detection comprises: acquiring feature maps output by convolutional layers in a Backbone network; performing convolution on the feature maps to obtain input feature maps of feature layers, the feature layers representing resolutions of the input feature maps; and fusing, based on densely connection feature pyramid network features, the input feature maps of each feature layer to obtain output feature maps of the feature layer. Since no additional convolutional layer is introduced for feature fusion, the detection performance for small targets may be enhanced without additional parameters, and the detection ability for small targets may be improved with computing resource constraints.Type: ApplicationFiled: March 26, 2021Publication date: July 22, 2021Applicant: BEIJING BAIDU NETCOM SCIENCE AND TECHNOLOGY CO., LTD.Inventors: Binghong Wu, Yehui Yang, Yanwu Xu, Lei Wang
-
Publication number: 20210192728Abstract: The present application discloses an image processing method, an apparatus, an electronic device and a storage medium. A specific implementation is: acquiring an image to be processed; acquiring a grading array according to the image to be processed and a grading network model, where the grading network model is a model pre-trained according to mixed samples, the number of elements contained in the grading array is C?1, C is the number of lesion grades, C lesion grades include one lesion grade without lesion and C?1 lesion grades with lesion, and a kth element in the grading array is a probability of a lesion grade corresponding to the image to be processed being greater than or equal to a kth lesion grade, where 1?k?C?1, and k is an integer; determining the lesion grade corresponding to the image to be processed according to the grading array.Type: ApplicationFiled: November 13, 2020Publication date: June 24, 2021Inventors: Fangxin SHANG, Yehui YANG, Lei WANG, Yanwu XU
-
Publication number: 20200320686Abstract: Embodiments of the present disclosure provide a method and apparatus for processing a fundus image. The method may include: acquiring a target fundus image; dividing the target fundus image into at least two first image blocks; inputting a first image block into a pre-trained deep learning model, to obtain a first output value; and determining, based on the first output value and a threshold, whether the first image block is the fundus image block containing a predetermined type of image region.Type: ApplicationFiled: December 2, 2019Publication date: October 8, 2020Inventors: Yehui Yang, Yanwu Xu, Lei Wang, Yan Huang
-
Publication number: 20200260944Abstract: A method and a device for recognizing a macular region and a computer-readable storage medium are provided. The method includes: obtaining a fundus image of a target object; extracting blood vessel information and optic disc information from the fundus image; inputting the blood vessel information and the optic disc information into a regression model to obtain location information of a macular fovea; and determining location information of the macular region of an eye of the target object, based on the location information of the macula fovea. In the embodiments of the application, the problem that the macular region cannot be accurately recognized when the image quality of the macular region is impaired is solved.Type: ApplicationFiled: November 27, 2019Publication date: August 20, 2020Applicant: Baidu Online Network Technology (Beijing) Co., Ltd.Inventors: Qinpei SUN, Yehui YANG, Lei WANG, Yanwu XU, Yan HUANG
-
Publication number: 20200218582Abstract: A device and method for automatically allocating computing resources is disclosed herein.Type: ApplicationFiled: March 18, 2020Publication date: July 9, 2020Inventors: Lanlan CONG, Heshan LIN, Yehui YANG
-
Patent number: 10650236Abstract: A road detection method and apparatus. A specific embodiment of the method includes: acquiring an image of a predetermined region; semantically segmenting the image to acquire a first probability that a region corresponding to each pixel in the image is a road region; acquiring a historical position information set of a target terminal; correcting, in response to historical position information existing in the historical position information set, the historical position information indicating a historical position located in the predetermined region, the first probability according to the historical position information to obtain a second probability; and determining a region corresponding to a pixel having the second probability greater than a preset threshold as a road region. Such an embodiment improves the road detection accuracy.Type: GrantFiled: January 30, 2018Date of Patent: May 12, 2020Assignee: Baidu Online Network Technology (Beijing) Co., Ltd.Inventors: Yuan Xia, Yehui Yang, Haishan Wu, Jingbo Zhou, Chao Li
-
Patent number: 10606662Abstract: A device and method for automatically allocating computing resources is disclosed herein.Type: GrantFiled: September 20, 2016Date of Patent: March 31, 2020Assignee: ALIBABA GROUP HOLDING LIMITEDInventors: Lanlan Cong, Heshan Lin, Yehui Yang
-
Patent number: 10572754Abstract: The present disclosure provides an area of interest boundary extracting method and apparatus, a device and a computer storage medium, wherein the area of interest boundary extracting method comprises: obtaining a satellite image and a road network base map including an area of interest; merging the obtained satellite image and road network base map to obtain merged data; using a binarizing model to perform binarization for the merged data to obtain a binarized image, wherein the binarizing model is obtained by training according to training data in advance; extracting a boundary of the binarized image as the boundary of the area of interest.Type: GrantFiled: March 2, 2018Date of Patent: February 25, 2020Assignee: BAIDU ONLINE NETWORK TECHNOLOGY (BEIJING) CO., LTD.Inventors: Yehui Yang, Yuan Xia, Haishan Wu, Chao Li
-
Patent number: 10382551Abstract: Apparatuses and methods are disclosed for cloud file processing. An exemplary method may include acquiring a position of a first data packet read by the client in the target file as a reading position when a target file that a client intends to read from a cloud server is divided into a plurality of data packets. The method may also include acquiring a state of a second data packet not at the reading position in the target file as a peripheral state. The peripheral state is a read-completed state or an unread state. When the reading position and the peripheral state meet a preset condition, the method may further include sending, by the client, a data pre-fetch request to the cloud server. The data pre-fetch request is used to request to read, from the cloud server, a preset amount of data packets whose peripheral states are unread states in the target file.Type: GrantFiled: October 28, 2016Date of Patent: August 13, 2019Assignee: ALIBABA GROUP HOLDING LIMITEDInventor: Yehui Yang
-
Publication number: 20180300549Abstract: A road detection method and apparatus. A specific embodiment of the method includes: acquiring an image of a predetermined region; semantically segmenting the image to acquire a first probability that a region corresponding to each pixel in the image is a road region; acquiring a historical position information set of a target terminal; correcting, in response to historical position information existing in the historical position information set, the historical position information indicating a historical position located in the predetermined region, the first probability according to the historical position information to obtain a second probability; and determining a region corresponding to a pixel having the second probability greater than a preset threshold as a road region. Such an embodiment improves the road detection accuracy.Type: ApplicationFiled: January 30, 2018Publication date: October 18, 2018Inventors: Yuan Xia, Yehui Yang, Haishan Wu, Jingbo Zhou, Chao Li
-
Publication number: 20180260648Abstract: The present disclosure provides an area of interest boundary extracting method and apparatus, a device and a computer storage medium, wherein the area of interest boundary extracting method comprises: obtaining a satellite image and a road network base map including an area of interest; merging the obtained satellite image and road network base map to obtain merged data; using a binarizing model to perform binarization for the merged data to obtain a binarized image, wherein the binarizing model is obtained by training according to training data in advance; extracting a boundary of the binarized image as the boundary of the area of interest.Type: ApplicationFiled: March 2, 2018Publication date: September 13, 2018Applicant: BAIDU ONLINE NETWORK TECHNOLOGY (BEIJING) CO., LTDInventors: Yehui YANG, Yuan Xia, Haishan Wu, Chao Li
-
Publication number: 20170192979Abstract: Methods and devices for accessing a cloud storage service based on a traditional file system interface. In one implementation, the method may include acquiring a traditional file access request sent by a client application; determining whether the traditional file access request is related to a cloud storage service system based on pre-stored file or disk information related to the cloud storage service system; responsive to determining that the traditional file access request is related to a cloud storage service system, converting the traditional file access request into an access request recognizable by the cloud storage service system and initiating an access to the cloud storage service system; and receiving result data returned by the cloud storage service system, converting the result data into a traditional file format, and returning the result data to the client application.Type: ApplicationFiled: December 29, 2016Publication date: July 6, 2017Inventor: Yehui YANG