Patents by Inventor Hsiang-Tsung Kung

Hsiang-Tsung Kung 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: 10586044
    Abstract: An abnormal behavior detection model building apparatus and an abnormal behavior detection model building method thereof are provided. The abnormal behavior detection model building apparatus analyzes the parts of speech of a plurality of program operation sequences in a plurality of program operation sequence data associated with abnormal behaviors to generate a plurality of word vectors and cluster the word vectors. Based on the result of the clustering, the abnormal behavior detection model building apparatus obtains a feature vector of each of the program operation sequence data, and perform a supervised learning for a classification algorithm by using the feature vectors so as to build an abnormal behavior detection model.
    Type: Grant
    Filed: December 12, 2017
    Date of Patent: March 10, 2020
    Assignee: Institute For Information Industry
    Inventors: Te-EN Wei, Chih-Hung Hsieh, Hsiang-Tsung Kung
  • Patent number: 10387797
    Abstract: A processor includes a front end to decode an instruction, an allocator to pass the instruction to a nearest neighbor logic unit (NNLU) to execute the instruction, and a retirement unit to retire the instruction. The NNLU includes logic to determine input of the instruction for which nearest neighbors will be calculated, transform the input, retrieve candidate atoms for which the nearest neighbors will be calculated, compute distance between the candidate atoms and the input, and determine the nearest neighbors for the input based upon the computed distance.
    Type: Grant
    Filed: September 25, 2015
    Date of Patent: August 20, 2019
    Assignee: Intel Corporation
    Inventors: Tsung-Han Lin, Gokce Keskin, Hsiang-Tsung Kung, She-Hwa Yen, Hong Wang
  • Publication number: 20190180030
    Abstract: An abnormal behavior detection model building apparatus and an abnormal behavior detection model building method thereof are provided. The abnormal behavior detection model building apparatus analyzes the parts of speech of a plurality of program operation sequences in a plurality of program operation sequence data associated with abnormal behaviors to generate a plurality of word vectors and cluster the word vectors. Based on the result of the clustering, the abnormal behavior detection model building apparatus obtains a feature vector of each of the program operation sequence data, and perform a supervised learning for a classification algorithm by using the feature vectors so as to build an abnormal behavior detection model.
    Type: Application
    Filed: December 12, 2017
    Publication date: June 13, 2019
    Inventors: Te-EN WEI, Chih-Hung HSIEH, Hsiang-Tsung KUNG
  • Patent number: 10296660
    Abstract: Detailed herein are embodiments of systems, methods, and apparatuses to be used for feature searching using an entry-based searching structure.
    Type: Grant
    Filed: April 21, 2014
    Date of Patent: May 21, 2019
    Assignee: INTEL CORPORATION
    Inventor: Hsiang-Tsung Kung
  • Patent number: 10282465
    Abstract: Detailed herein are embodiments of systems, methods, and apparatuses to be used for feature searching using an entry-based searching structure.
    Type: Grant
    Filed: June 20, 2014
    Date of Patent: May 7, 2019
    Assignee: Intel Corporation
    Inventors: Tsung-Han Lin, Hsiang-Tsung Kung
  • Patent number: 10262270
    Abstract: A system and a method for predicting a remaining useful life (RUL) of a component of an equipment are provided. The system for predicting the RUL of the component of the equipment includes a data acquisition unit, a feature capturing unit, a mapping function generating unit, a similarity analyzing unit and a RUL calculating unit. The feature capturing unit obtains an estimation feature according to a real time sensing record, and obtains a plurality of training features according to a set of history sensing records. The similarity analyzing unit obtains k similar features which are similar to the estimation feature according to the training features. The RUL calculating unit obtains at least one of k predicting information via a mapping function according to the k similar features and calculates an estimation RUL according to at least one predicting value.
    Type: Grant
    Filed: August 17, 2016
    Date of Patent: April 16, 2019
    Assignee: INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE
    Inventors: Hsiang-Tsung Kung, Jia-Min Ren, Chuang-Hua Chueh, Sen-Chia Chang
  • Publication number: 20190095787
    Abstract: System and techniques for sparse coding based classification are described herein. A sample of a first type of data may be obtained and encoded to create a sparse coded sample. A dataset may be searched using the sparse coded sample to locate a segment set of a second type of data. An instance of the second type of data may then be created using the segment set.
    Type: Application
    Filed: September 27, 2017
    Publication date: March 28, 2019
    Inventors: Hsiang Tsung Kung, Chit Kwan Lin, Gautham N. Chinya
  • Patent number: 9779619
    Abstract: Methods and systems for operating a sensor network comprising a plurality of nodes and at least one resource, where a resource can be a sensing device (e.g., a camera or microphone) or other peripheral devices (e.g., a storage or recording system). In at least one embodiment of the present invention, at least one node in a network determines at least one resource available to it and the type of the resource or its capabilities. Based on the type or capabilities of the resource, the node associates with one or more logical node names and acquires instructions for performing at least one function.
    Type: Grant
    Filed: March 15, 2013
    Date of Patent: October 3, 2017
    Assignee: Industrial Technology Research Institute and Hsiang-Tsung Kung
    Inventors: Hsiang-Tsung Kung, Dario Vlah
  • Patent number: 9747788
    Abstract: Methods and systems for operating a sensor network comprising a plurality of nodes and at least one resource, where a resource can be a sensing device (e.g., a camera or microphone) or other peripheral devices (e.g., a storage or recording system). In at least one embodiment of the present invention, at least one node in a network determines at least one resource available to it and the type of the resource or its capabilities. Based on the type or capabilities of the resource, the node associates with one or more logical node names and acquires instructions for performing at least one function.
    Type: Grant
    Filed: March 15, 2013
    Date of Patent: August 29, 2017
    Assignee: Industrial Technology Research Institute
    Inventors: Hsiang-Tsung Kung, Dario Vlah
  • Publication number: 20170185900
    Abstract: An apparatus is described herein. The apparatus includes a clustering mechanism that is to partition a dictionary into a plurality of clusters. The apparatus also includes a feature-matching mechanism that is to pre-compute feature matching results for each cluster of the plurality of clusters. Moreover, the apparatus includes a selector that is to locate a best representative feature from the dictionary in response to an input vector.
    Type: Application
    Filed: December 26, 2015
    Publication date: June 29, 2017
    Applicant: INTEL CORPORATION
    Inventors: Arnab Paul, Nilesh Jain, Hsiang-Tsung Kung
  • Publication number: 20170161622
    Abstract: A system and a method for predicting a remaining useful life (RUL) of a component of an equipment are provided. The system for predicting the RUL of the component of the equipment includes a data acquisition unit, a feature capturing unit, a mapping function generating unit, a similarity analyzing unit and a RUL calculating unit. The feature capturing unit obtains an estimation feature according to a real time sensing record, and obtains a plurality of training features according to a set of history sensing records. The similarity analyzing unit obtains k similar features which are similar to the estimation feature according to the training features. The RUL calculating unit obtains at least one of k predicting information via a mapping function according to the k similar features and calculates an estimation RUL according to at least one predicting value.
    Type: Application
    Filed: August 17, 2016
    Publication date: June 8, 2017
    Applicant: INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE
    Inventors: Hsiang-Tsung KUNG, Jia-Min REN, Chuang-Hua CHUEH, Sen-Chia CHANG
  • Patent number: 9626862
    Abstract: Methods and systems for operating a sensor network comprising a plurality of nodes and at least one resource, where a resource can be a sensing device (e.g., a camera or microphone) or other peripheral devices (e.g., a storage or recording system). In at least one embodiment of the present invention, at least one node in a network determines at least one resource available to it and the type of the resource or its capabilities. Based on the type or capabilities of the resource, the node associates with one or more logical node names and acquires instructions for performing at least one function.
    Type: Grant
    Filed: June 18, 2004
    Date of Patent: April 18, 2017
    Assignee: Industrial Technology Research Institute
    Inventors: Hsiang-Tsung Kung, Dario Vlah
  • Publication number: 20170091655
    Abstract: A processor includes a front end to decode an instruction, an allocator to pass the instruction to a nearest neighbor logic unit (NNLU) to execute the instruction, and a retirement unit to retire the instruction. The NNLU includes logic to determine input of the instruction for which nearest neighbors will be calculated, transform the input, retrieve candidate atoms for which the nearest neighbors will be calculated, compute distance between the candidate atoms and the input, and determine the nearest neighbors for the input based upon the computed distance.
    Type: Application
    Filed: September 25, 2015
    Publication date: March 30, 2017
    Inventors: Tsung-Han Lin, Gokce Keskin, Hsiang-Tsung Kung, She-Hwa Yen, Hong Wang
  • Patent number: 9606167
    Abstract: Circuitry, systems and methods for testing integrated circuits for the presence of anomalies. Techniques include applying a plurality of inputs to an integrated circuit under test to obtain a first plurality of measurements at least partially characterizing power leakage in the integrated circuit under test, encode the first plurality of measurements, by computing a plurality of random linear combinations of measurements in the first plurality of measurements, to obtain a second plurality of encoded measurements determining whether the integrated circuit under test contains at least one anomaly based, at least in part, on the second plurality of encoded measurements.
    Type: Grant
    Filed: August 2, 2012
    Date of Patent: March 28, 2017
    Assignee: President and Fellows of Harvard College
    Inventors: Hsiang-Tsung Kung, Dario Vlah
  • Patent number: 9178590
    Abstract: A channel information feedback method for multi-antenna system, and a wireless communication device using the same method are provided. The proposed method could reduce feedback overhead for multiple-input multiple-output (MIMO) wireless channel information, and is based on compressive sensing technique. Prior to sending back channel information, a receiver estimates the channel and multiplies the vectorized channel with a random matrix to generate compressed feedback content. Upon receiving the compressed feedback content at a transmitter, the channel information could be restored with signal recovery algorithms of compressive sensing technique. In the other embodiment, the proposed method further adaptively adjusts compression ratio of the compressed feedback content in accordance to the prevailing channel quality. Further, for slow-varying MIMO channels, there is proposed another channel information feedback method which switches between a fixed sparcifying-basis and a signal-dependent sparcifying-basis.
    Type: Grant
    Filed: December 27, 2011
    Date of Patent: November 3, 2015
    Assignee: Industrial Technology Research Institute
    Inventors: Ping-Heng Kuo, Hsiang-Tsung Kung, Pang-An Ting
  • Publication number: 20150242463
    Abstract: Detailed herein are embodiments of systems, methods, and apparatuses to be used for feature searching using an entry-based searching structure.
    Type: Application
    Filed: June 20, 2014
    Publication date: August 27, 2015
    Inventors: Tsung-Han Lin, Hsiang-Tsung Kung
  • Publication number: 20150242461
    Abstract: Detailed herein are embodiments of systems, methods, and apparatuses to be used for feature searching using an entry-based searching structure.
    Type: Application
    Filed: April 21, 2014
    Publication date: August 27, 2015
    Inventor: Hsiang-Tsung Kung
  • Publication number: 20150234908
    Abstract: Techniques for data synchronization between a first computing device coupled to at least one memory storing current data and a second computing device coupled to at least a second memory storing first encoded data and a copy of prior data. The first device may perform a method comprising: encoding the current data using a compressive sensing encoding technique to obtain second encoded data; and transmitting the second encoded data to the second computing device. The second device may perform a method comprising receiving second encoded data from the first computing device; decoding the second encoded data using a compressive sensing decoding technique to obtain decoded data; and obtaining a copy of the current data by using the decoded data and the copy of prior data.
    Type: Application
    Filed: September 24, 2013
    Publication date: August 20, 2015
    Applicant: President and Fellows of Harvard College
    Inventors: Hsiang-Tsung Kung, Chit-Kwan Lin
  • Patent number: 9032075
    Abstract: Techniques for medium access control. Some techniques include receiving, at a first computing device, a solicitation for at least a first medium access request that specifies at least one time period for transmitting the first medium access request to the second computing device; encoding the first medium access request at least in part by using a compressive sensing encoding technique to obtain a first encoded medium access request; and transmitting the first encoded medium access request to the second computing device during the at least one time period specified in the received solicitation.
    Type: Grant
    Filed: October 11, 2012
    Date of Patent: May 12, 2015
    Assignee: President and Fellows of Harvard College
    Inventors: Hsiang-Tsung Kung, Tsung-Han Lin
  • Patent number: 8957886
    Abstract: A context-showing and power-saving display information converting system and a corresponding apparatus are provided for lowering the power consumption of displaying images on a pixel self-emissive display device. The display information converting system includes a context-showing module and a power-saving conversion module. The context-showing module receives display information that describes the content of an image, wherein the display information includes one or more elements, and each element corresponds to a plurality of pixels displayed by a display device. In addition, the context-showing module determines a relevance of the one or more elements according to the viewing interest of the user. The power-saving conversion module converts the display information in unit of the one or more elements according to the relevance and provides the converted display information to the display device.
    Type: Grant
    Filed: January 13, 2011
    Date of Patent: February 17, 2015
    Assignee: Industrial Technology Research Institute
    Inventors: Tun-Hao You, Yi-Jui Wu, Hsiang-Tsung Kung, Koan-Sin Tan, Chen-Hwa Song, Yi-Jen Yeh