Patents by Inventor Shiwei HUANG
Shiwei HUANG 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).
-
Publication number: 20250094792Abstract: A task execution method for a large model, an electronic device, and a storage medium are provided, which relate to a field of artificial intelligence technology, particularly to fields of deep learning technology and large model technology.Type: ApplicationFiled: December 4, 2024Publication date: March 20, 2025Inventors: Bo KE, Xuyi CHEN, Zhengjie HUANG, Shikun FENG, Weibin LI, Shiwei HUANG
-
Patent number: 12229519Abstract: A method for generating a dialogue state includes: acquiring a target dialogue state of a previous round of dialogue and dialogue information of a current round of dialogue; generating an initial dialogue state of the current round of dialogue according to the target dialogue state of the previous round of dialogue and the dialogue information of the current round of dialogue; and generating a target dialogue state of the current round of dialogue according to the initial dialogue state of the current round of dialogue and the dialogue information of the current round of dialogue.Type: GrantFiled: June 9, 2022Date of Patent: February 18, 2025Assignee: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.Inventors: Xin Tian, Liankai Huang, Yingzhan Lin, Siqi Bao, Huang He, Fan Wang, Shuqi Sun, Shiwei Huang
-
Publication number: 20250032446Abstract: A novel yeast strain and methods of use thereof for robust biosynthesis of demethylated polymethoxy flavones are disclosed. Also provided are method of administering demethylated polymethoxy flavones for the treatment of obesity and inflammatory a disease.Type: ApplicationFiled: December 12, 2022Publication date: January 30, 2025Applicant: RUTGERS, THE STATE UNIVERSITY OF NEW JERSEYInventors: Shiwei SU, Chi-Tang HO, Qingrong HUANG
-
Publication number: 20250028958Abstract: A data processing method, and a data processing model and a training method therefor are provided, and relate to the field of artificial intelligence, and specifically, to natural language processing, deep learning technologies, and large model technologies. An implementation solution includes: determining input data, where the input data includes a plurality of tokens; determining a correlation between each of the plurality of tokens and each of a plurality of expert networks based on a gating matrix, where the plurality of expert networks are used to reinforce the plurality of tokens; allocating the plurality of tokens to the plurality of expert networks in a uniform manner based on the correlation and a preset capacity of each expert network, to reinforce the plurality of tokens; and determining a data processing result based on the plurality of reinforced tokens.Type: ApplicationFiled: October 7, 2024Publication date: January 23, 2025Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.Inventors: Xuyi CHEN, Bo KE, Chenhui LI, Zhengjie HUANG, Shiwei HUANG, Weibin LI, Shikun FENG
-
Publication number: 20240388274Abstract: A bonding structure and an acoustic wave device relate to the field of filters. The bonding structure includes a supporting substrate and a piezoelectric layer formed on the supporting substrate. The supporting substrate is made of a polycrystalline material, and a porosity of the supporting substrate is less than 0.0045% or greater than 0.6%. The bonding structure can effectively improve the generation of interference of spurious emission and improve the performance of the device.Type: ApplicationFiled: July 30, 2024Publication date: November 21, 2024Inventors: Zhonghe LIN, Minghui FANG, Yenfu LIN, Shiwei HUANG, Shengyu YANG
-
Patent number: 11775766Abstract: Embodiments of a method and an apparatus for improving a model based on a pre-trained semantic model are provided. The method may include: based on the pre-trained semantic model, obtaining an initial improved model, where semantic result information of an input vector is determined in the initial improved model based on a hash search method; and based on a model distillation method, training the initial improved model to obtain an improved model. Some embodiments can obtain the semantic result information of the input vector by performing the hash search method on the input vector, replace the original complex iterative calculation process of a semantic model, and obtain the improved model with few model parameters and high compression ratio.Type: GrantFiled: March 10, 2021Date of Patent: October 3, 2023Assignee: BEIJING BAIDU NETCOM SCIENCE AND TECHNOLOGY CO., LTD.Inventors: Xuyi Chen, Shiwei Huang
-
Publication number: 20230222827Abstract: In a method for processing a document image, a document image to be processed is acquired. Text nodes of multiple granularities, visual nodes of multiple granularities, respective node information of the text nodes, and respective node information of the visual nodes in the document image are obtained. A multi-granularity and multi-modality document graph is construct based on the text nodes of multiple granularities, the visual nodes of multiple granularities, the respective node information of the text nodes and the respective node information of the visual nodes. Multi-granularity semantic feature information of the document image is determined based on the multi-granularity and multi-modality document graph, the respective node information of the text nodes and the respective node information of the visual nodes.Type: ApplicationFiled: March 10, 2023Publication date: July 13, 2023Inventors: Wenjin Wang, Zhengjie Huang, Bin Luo, Qiming Peng, Weichong Yin, Shikun Feng, Shiwei Huang, Jingzhou He
-
Publication number: 20230214689Abstract: A method for processing a dialogue includes: obtaining a dialogue text of the dialogue, in which the dialogue text includes a current question text, or the dialogue text includes the current question text and a historical dialogue text; extracting a current query text from the dialogue text; obtaining a knowledge query result for the current query text by querying a knowledge database based on the current query text; and determining a response text for the current question text based on the knowledge query result and the dialogue text.Type: ApplicationFiled: March 14, 2023Publication date: July 6, 2023Inventors: Xin TIAN, Yingzhan LIN, Mengfei SONG, Siqi BAO, Shiwei HUANG
-
Publication number: 20230141932Abstract: A method for answer questioning based on a table includes the following. A question text to be processed and an information table for question answering are determined, and the information table includes: at least one attribute name. A character vector sequence, a position vector sequence and a type vector sequence are determined based on the question text and the at least one attribute name. An attribute name segment and an attribute value segment in the question text are determined based on the character vector sequence, the position vector sequence and the type vector sequence. An answer corresponding to the question text is determined based on the attribute name segment, the attribute value segment and the information table.Type: ApplicationFiled: December 1, 2022Publication date: May 11, 2023Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.Inventors: Dongfeng He, Bingjin Chen, Jiayang Tu, Yingzhan Lin, Shiwei Huang
-
Publication number: 20230085458Abstract: A method for generating dialog data is provided. An implementation is: obtaining a target dialog data template, where the target dialog data template includes one or more target single-round dialog data templates, each target single-round dialog data template includes one or more keyword slots and related information about each keyword slot, and the related information about each keyword slot includes location information and attribute information; for each keyword slot, determining, from a keyword data set at least based on the attribute information of the keyword slot, one or more target keywords that match the keyword slot; and for each target single-round dialog data template, filling the target single-round dialog data template with the one or more target keywords based on the location information of the one or more keyword slots, to obtain target dialog data.Type: ApplicationFiled: November 21, 2022Publication date: March 16, 2023Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.Inventors: Xin TIAN, Dongfeng HE, Liankai HUANG, Yingzhan LIN, Shiwei HUANG
-
Publication number: 20230047980Abstract: A method of training a deep learning model, a method of processing a natural language, an electronic device, and a storage medium are provided, which relate to a field of artificial intelligence, in particular to deep learning technology and natural language processing technology. The method includes: inputting first sample data into a first deep learning model to obtain a first output result; training the first deep learning model according to the first output result and a first target output result, the first target output result is obtained by processing the first sample data using a reference deep learning model; inputting second sample data into a second deep learning model to obtain a second output result; and training the second deep learning model according to the second output result and a second target output result, to obtain a trained second deep learning model.Type: ApplicationFiled: October 28, 2022Publication date: February 16, 2023Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.Inventors: Xuyi CHEN, Weixin LIU, Yuxiang LU, Jiaxiang LU, Shiwei HUANG
-
Publication number: 20230004774Abstract: The present disclosure provides a method and apparatus for generating a node representation, an electronic device and a readable storage medium, and relates to the field of deep learning technologies. The method for generating a node representation includes: acquiring a heterogeneous graph to be processed; performing a sampling operation in the heterogeneous graph to be processed according to a first meta path, so as to obtain at least one first walk path; obtaining an initial node representation of each node in the heterogeneous graph to be processed according to the at least one first walk path; and generating the final node representation of each node according to the initial node representation of each node and initial node representations of neighbor nodes of each node. With the present disclosure, accuracy of the generated node representation may be improved.Type: ApplicationFiled: January 19, 2022Publication date: January 5, 2023Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.Inventors: Weibin LI, Zhifan ZHU, Shikun FENG, Shiwei HUANG, Jingzhou HE
-
Publication number: 20220300717Abstract: A method for generating a dialogue state includes: acquiring a target dialogue state of a previous round of dialogue and dialogue information of a current round of dialogue; generating an initial dialogue state of the current round of dialogue according to the target dialogue state of the previous round of dialogue and the dialogue information of the current round of dialogue; and generating a target dialogue state of the current round of dialogue according to the initial dialogue state of the current round of dialogue and the dialogue information of the current round of dialogue.Type: ApplicationFiled: June 9, 2022Publication date: September 22, 2022Applicant: BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD.Inventors: Xin Tian, Liankai Huang, Yingzhan Lin, Siqi Bao, Huang He, Fan Wang, Shuqi Sun, Shiwei Huang
-
Patent number: 11397852Abstract: A news interaction method, apparatus, device and computer storage medium are proposed. Input information input by a user upon reading current news content is obtained; parsing information of the input information is obtained based on the current news content, where the parsing information includes intent information of the input information; the input information is distributed to at least one news interactive service subsystem according to the intent information of the input information, and a return result returned by the at least one news interactive service subsystem is received; and a display result is selected from the return result according to a preset policy, and provided to the user.Type: GrantFiled: December 13, 2019Date of Patent: July 26, 2022Assignee: BEIJING BAIDU NETCOM SCIENCE AND TECHNOLOGY CO., LTDInventors: Shuo Huang, Jiaxin Lin, Zhihong Fu, Jinbo Zhan, Guang Ling, Shiwei Huang, Guyue Zhou, Chao Zhou
-
Publication number: 20220129753Abstract: A pre-training method of a neural network model, an electronic device, and a medium. The pre-training data is inputted to the initial neural network model, and the initial neural network model is pre-trained in the first training mode, in the first training mode, the plurality of hidden layers share one hidden layer parameter, and the loss value of the initial neural network model is obtained, if the loss value of the initial neural network model is less than a preset threshold, the initial neural network model continues to be pre-trained in the second training mode, in the second training mode, each of the plurality of hidden layers has its own hidden layer parameter.Type: ApplicationFiled: January 11, 2022Publication date: April 28, 2022Inventors: Yuxiang LU, Jiaxiang LIU, Xuyi CHEN, Shikun FENG, Shuohuan WANG, Yu SUN, Shiwei HUANG, Jingzhou HE
-
Publication number: 20220129448Abstract: An intelligent dialogue method and apparatus and medium are provided. The method includes: obtaining a pre-matching result by pre-matching a query to be processed with a table content of a target table; extracting a character segment having a highest matching degree with the attribute value from the query based on the attribute value having the highest matching degree with the query; determining a target attribute value semantically associated with the character segment based on the attribute value of each column attribute; generating a structured query language (SQL) query statement corresponding to the query based on the query, the attribute name of each column attribute, the highest matching level of the attribute name, the highest matching level of the attribute value and the target attribute value; and generating a reply statement based on a result obtained by searching a database based on the SQL query statement.Type: ApplicationFiled: January 6, 2022Publication date: April 28, 2022Applicant: Beijing Baidu Netcom Science Technology Co., Ltd.Inventors: Dongfeng He, Bingjin Chen, Wanshun Chen, Jiayang Tu, Yingzhan Lin, Shiwei Huang
-
Patent number: 11208124Abstract: The present application relate to a follow-up mechanism and a brake caliper unit for gauge-changeable bogie, the follow-up mechanism includes a follow-up connector, unlocking members that are located on two sides of the follow-up connector and movably connected to the follow-up connector, a transverse displacement recognition device movably connected to the unlocking members, a toothed locking and positioning device mounted on the follow-up connector, and at least two mutually parallel fixation members; the follow-up connector is in sliding fit with the fixation members, and sliders are fixedly connected at ends of the unlocking members; the toothed locking and positioning device is movably connected to the transverse displacement recognition device and fixation members, respectively; the brake caliper unit comprises a mounting bracket, the follow-up mechanism, and a brake actuator mounted on the mounting bracket, the follow-up mechanism is installed in cooperation with the brake actuator, and the fixation meType: GrantFiled: May 18, 2020Date of Patent: December 28, 2021Assignee: CRRC QINGDAO SIFANG ROLLING STOCK RESEARCH INSTITUTE CO., LTD. (CN)Inventors: Zhen Wang, Qingyu Meng, Lingjun Wang, Xiaochao Dai, Xin Zhang, Fengzhou Wang, Jiansong Huang, Fangliang Zhang, Shiwei Huang
-
Publication number: 20210397947Abstract: Embodiments of the present disclosure provide a method for generating a model for representing heterogeneous graph node. A specific implementation includes: acquiring a training data set, wherein the training data set includes node walk path information obtained by sampling a heterogeneous graph according to different meta paths; and training, based on a gradient descent algorithm, an initial heterogeneous graph node representation model with the training data set as an input of the initial heterogeneous graph node representation model, to obtain a heterogeneous graph node representation model.Type: ApplicationFiled: December 9, 2020Publication date: December 23, 2021Inventors: Weibin LI, Zhifan ZHU, Shikun FENG, Jingzhou HE, Shiwei HUANG
-
Publication number: 20210397794Abstract: Embodiments of a method and an apparatus for improving a model based on a pre-trained semantic model are provided. The method may include: based on the pre-trained semantic model, obtaining an initial improved model, where semantic result information of an input vector is determined in the initial improved model based on a hash search method; and based on a model distillation method, training the initial improved model to obtain an improved model. Some embodiments can obtain the semantic result information of the input vector by performing the hash search method on the input vector, replace the original complex iterative calculation process of a semantic model, and obtain the improved model with few model parameters and high compression ratio.Type: ApplicationFiled: March 10, 2021Publication date: December 23, 2021Inventors: Xuyi CHEN, Shiwei HUANG
-
Publication number: 20210383233Abstract: The disclosure discloses a method for distilling a model, an electronic device, and a storage medium, and relates to the field of deep learning technologies. A teacher model and a student model are obtained. The second intermediate fully connected layer is transformed into an enlarged fully connected layer and a reduced fully connected layer based on a first data processing capacity of a first intermediate fully connected layer of the teacher model and a second data processing capacity of a second intermediate fully connected layer of the student model. The second intermediate fully connected layer is replaced with the enlarged fully connected layer and the reduced fully connected layer to generate a training student model. The training student model is distilled based on the teacher model.Type: ApplicationFiled: November 23, 2020Publication date: December 9, 2021Inventors: Weiyue SU, Shikun FENG, Zhifan ZHU, Weibin LI, Jingzhou HE, Shiwei HUANG