Patents by Inventor Wan-Yi Lin

Wan-Yi Lin 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: 20210042530
    Abstract: A storage maintains raw image data including video having a sequence of frames, and annotations of the frames that indicate aspects of objects identified in the respective frames. A processor determines, for annotation of key frames of the raw image data, a task type for key frames and an agent type for key frames, receives annotations of objects identified in the key frames of the raw image data according to the key frame task type and key frame agent type, selects to review the key frames based on a confidence level of the annotations of the key frames, determines, for annotation of intermediate frames of the raw image data, a task type for intermediate frames and an agent type for intermediate frames, and receives annotations of objects identified in the intermediate frames of the raw image data according to the intermediate frame task type and intermediate frame agent type.
    Type: Application
    Filed: August 8, 2019
    Publication date: February 11, 2021
    Inventors: Ji Eun KIM, Wan-Yi LIN, Lixiu YU, Govind RATHORE
  • Publication number: 20200380258
    Abstract: A data annotation system for a machine-learning system includes a computing system configured to execute a machine-learning algorithm for generating data annotations of input data. The computing system is further configured to execute an interface for communicating results and predictions of the machine-learning algorithm to non-expert workers for review. The interface communicates with the non-expert worker using visualization and natural language to elicit feedback to improve performance of the machine-learning algorithm.
    Type: Application
    Filed: May 31, 2019
    Publication date: December 3, 2020
    Inventors: Lixiu YU, Wan-yi LIN, Ji Eun KIM
  • Publication number: 20200210855
    Abstract: An information synthesis system for generating a knowledge base injects domain knowledge into semi-crowdsourced summarization pipelines for extracting information from unstructured data sources. The summarization pipeline includes chains of tasks completed by crowd workers and/or machined. The information synthesis system distributes the tasks to crowd workers and/or machines. Task responses are processed and aggregated to determine new information that is used to update the knowledge base.
    Type: Application
    Filed: December 28, 2018
    Publication date: July 2, 2020
    Inventors: Ji Eun KIM, Wan-Yi LIN, Lixiu YU
  • Publication number: 20200209000
    Abstract: A decision-support system for a public transportation system includes a computing system programmed to generate public transit route options for the commuter based on at least one decision factor that estimates an expected impact on other commuters resulting from the commuter choosing each of the public transit route options, and output, for display, values associated with the at least one decision factor to influence the commuter in making a route selection.
    Type: Application
    Filed: December 28, 2018
    Publication date: July 2, 2020
    Inventors: Alessandro OLTRAMARI, Wan-Yi LIN, Lixiu YU
  • Patent number: 10701261
    Abstract: A method includes analyzing a preview image made by an image sensor when operating in a preview image mode to determine if the preview image contains some portion that corresponds to information that is considered not subject to being imaged. The method further includes, if it is determined that the preview image does comprise some image portion that corresponds to information that is considered not subject to being imaged, disabling the image sensor from capturing an image that would correspond to the preview image. Analyzing can include accessing a data storage that contains representations of information that is considered not subject to being imaged for comparison with image features extracted from the preview image, where the data storage is at least one of contained in a device that also contains the image sensor, or is located remotely from the device and can be accessed via a wireless connection.
    Type: Grant
    Filed: August 1, 2016
    Date of Patent: June 30, 2020
    Assignee: International Business Machines Corporation
    Inventors: Jui-Hsin Lai, Ching-Yung Lin, Wan-Yi Lin
  • Publication number: 20180035045
    Abstract: A method includes analyzing a preview image made by an image sensor when operating in a preview image mode to determine if the preview image contains some portion that corresponds to information that is considered not subject to being imaged. The method further includes, if it is determined that the preview image does comprise some image portion that corresponds to information that is considered not subject to being imaged, disabling the image sensor from capturing an image that would correspond to the preview image. Analyzing can include accessing a data storage that contains representations of information that is considered not subject to being imaged for comparison with image features extracted from the preview image, where the data storage is at least one of contained in a device that also contains the image sensor, or is located remotely from the device and can be accessed via a wireless connection.
    Type: Application
    Filed: August 1, 2016
    Publication date: February 1, 2018
    Inventors: Jui-Hsin Lai, Ching-Yung Lin, Wan-Yi Lin
  • Patent number: 9798982
    Abstract: Determining a number of kernels within a model is provided. A number of kernels that include data samples of a majority data class of an imbalanced training data set is determined based on a set of generated artificial data samples for a minority data class of the imbalanced training data set. The number of kernels within the model is generated based on the set of generated artificial data samples. A likelihood of the set of generated artificial data samples being included in the majority data class of the imbalanced training data set is calculated. Parameters of each kernel in the number of kernels are updated based on the likelihood of the set of generated artificial data samples being included in the majority data class of the imbalanced training data set. Each kernel in the number of kernels is adjusted based on the updated parameters.
    Type: Grant
    Filed: August 20, 2015
    Date of Patent: October 24, 2017
    Assignee: International Business Machines Corporation
    Inventors: Ching-Yung Lin, Wan-Yi Lin, Yinglong Xia
  • Patent number: 9589245
    Abstract: A method for predicting insider threat includes mining electronic data of an organization corresponding to activity of an entity, determining features of the electronic data corresponding to the activity of the entity, classifying the features corresponding to the activity of the entity, determining sequences of classified features matching one or more patterns of insider threat, scoring the entity according to matches of the classified features to the one or more patterns of insider threat, and predicting an insider threat corresponding to the entity according to the score.
    Type: Grant
    Filed: April 7, 2014
    Date of Patent: March 7, 2017
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Anni R. Coden, Ching-Yung Lin, Wan-Yi Lin, Yinglong Xia
  • Patent number: 9538040
    Abstract: Various communication systems may benefit from physical layer watermarking. For example, active sensing for dynamic spectrum access may be performed using physical layer watermarking, such as watermarking based on channel effects and/or receiver distortion. A method may include, for example, obtaining an original signal to be transmitted to at least one receiver. The method may also include watermarking the original signal with at least one of authentication data or ancillary data to provide an enhanced signal. The watermarking can include a physical layer watermark. The physical layer watermark can be configured to emulate at least one a channel effect or a receiver distortion. The method can further include transmitting the enhanced signal to the at least one receiver.
    Type: Grant
    Filed: December 17, 2012
    Date of Patent: January 3, 2017
    Assignee: UNIVERSITY OF MARYLAND, COLLEGE PARK
    Inventors: Nathan Goergen, Wan-Yi Lin, K. J. Ray Liu
  • Patent number: 9337973
    Abstract: The present disclosure proposes a method of cooperative MIMO wireless communication, and a base station using the same. A base station would have a plurality of radio remote units (RRUs), and each of the RRUs may have a specific radio coverage and a plurality of antennas coupled to each of the RRUs. The function of the base station may include receiving a UE information, receiving a service flow, deriving a service type of the service flow, and determining a MIMO scheme, modulating the service flow into a MIMO stream according to the MIMO scheme, allocating a plurality of resource blocks, wherein the resource blocks may correspond to a portion or all of the RRUs, distributing the MIMO stream into a plurality of downlink sub streams and transmitting the downlink sub streams through the resource blocks by the RRUs corresponding to the resource blocks.
    Type: Grant
    Filed: September 3, 2013
    Date of Patent: May 10, 2016
    Assignee: Industrial Technology Research Institute
    Inventors: Hsin-An Hou, Wan-Yi Lin, Yung-Han Chen
  • Patent number: 9224104
    Abstract: Injecting generated data samples into a minority data class of an imbalanced training data set is provided. In response to receiving an input to balance the imbalanced training data set that includes a majority data class and the minority data class, a set of data samples is generated for the minority data class. A distance is calculated from each data sample in the set of generated data samples to a center of a kernel that includes a set of data samples of the majority data class. Each data sample in the set of generated data samples is stored within a corresponding distance score bucket based on the calculated distance of a data sample. Generated data samples are selected from a number of highest ranking distance score buckets. The generated data samples selected from the number of highest ranking distance score buckets are injected into the minority data class.
    Type: Grant
    Filed: September 24, 2013
    Date of Patent: December 29, 2015
    Assignee: International Business Machines Corporation
    Inventors: Ching-Yung Lin, Wan-Yi Lin, Yinglong Xia
  • Publication number: 20150356464
    Abstract: Determining a number of kernels within a model is provided. A number of kernels that include data samples of a majority data class of an imbalanced training data set is determined based on a set of generated artificial data samples for a minority data class of the imbalanced training data set. The number of kernels within the model is generated based on the set of generated artificial data samples. A likelihood of the set of generated artificial data samples being included in the majority data class of the imbalanced training data set is calculated. Parameters of each kernel in the number of kernels are updated based on the likelihood of the set of generated artificial data samples being included in the majority data class of the imbalanced training data set. Each kernel in the number of kernels is adjusted based on the updated parameters.
    Type: Application
    Filed: August 20, 2015
    Publication date: December 10, 2015
    Inventors: Ching-Yung Lin, Wan-Yi Lin, Yinglong Xia
  • Publication number: 20150286819
    Abstract: A method for predicting insider threat includes mining electronic data of an organization corresponding to activity of an entity, determining features of the electronic data corresponding to the activity of the entity, classifying the features corresponding to the activity of the entity, determining sequences of classified features matching one or more patterns of insider threat, scoring the entity according to matches of the classified features to the one or more patterns of insider threat, and predicting an insider threat corresponding to the entity according to the score.
    Type: Application
    Filed: April 7, 2014
    Publication date: October 8, 2015
    Applicant: International Business Machines Corporation
    Inventors: Anni R. Coden, Ching-Yung Lin, Wan-Yi Lin, Yinglong Xia
  • Patent number: 8995706
    Abstract: A method and system of detecting colluders conducting a collusion attack including a minority-type collusion attack on a digital product includes the generation of codewords used as watermarks in the digital product. The inner code of the codewords is generated using permutations of rows in a Hadamard matrix and concatenating them together. A typical outer code of the codeword is the Reed Solomon code. An adaptive detector is able to accurately detect one of three or more colluders of a minority-type attack. Prior art schemes using an error correcting code-based watermarking mechanism with an inner code fail to detect colluders with a minority-type collusion attack which includes three colluders.
    Type: Grant
    Filed: September 26, 2008
    Date of Patent: March 31, 2015
    Assignee: Thomson Licensing
    Inventors: Wan-Yi Lin, Shan He, Jeffrey Adam Bloom
  • Publication number: 20150088791
    Abstract: Injecting generated data samples into a minority data class of an imbalanced training data set is provided. In response to receiving an input to balance the imbalanced training data set that includes a majority data class and the minority data class, a set of data samples is generated for the minority data class. A distance is calculated from each data sample in the set of generated data samples to a center of a kernel that includes a set of data samples of the majority data class. Each data sample in the set of generated data samples is stored within a corresponding distance score bucket based on the calculated distance of a data sample. Generated data samples are selected from a number of highest ranking distance score buckets. The generated data samples selected from the number of highest ranking distance score buckets are injected into the minority data class.
    Type: Application
    Filed: September 24, 2013
    Publication date: March 26, 2015
    Applicant: International Business Machines Corporation
    Inventors: Ching-Yung Lin, Wan-Yi Lin, Yinglong Xia
  • Patent number: 8745402
    Abstract: A method and apparatus are described including generating a unique code for each of a plurality of users using a plurality of symbols, generating a plurality of codes representing the plurality of symbols, substituting the plurality of codes into the unique code for each of the plurality of users, permuting the code resulting from the substitution to produce a codeword for each of the plurality of users and embedding the codeword into digital content. The second generating act further includes generating a string of first symbols followed by second symbols, wherein the first symbols are all ones and the second symbols are all negative ones, wherein a number of first symbols is equal to a number of the second symbols, and wherein if a length of the first symbols followed by the second symbols is less than a length of the code, then the first symbols followed by the second symbols are repeated until the code length is filled.
    Type: Grant
    Filed: September 26, 2008
    Date of Patent: June 3, 2014
    Assignee: Thomson Licensing
    Inventors: Wan-yi Lin, Shan He, Jeffrey Adam Bloom
  • Publication number: 20140071912
    Abstract: The present disclosure proposes a method of cooperative MIMO wireless communication, and a base station using the same. A base station would have a plurality of radio remote units (RRUs), and each of the RRUs may have a specific radio coverage and a plurality of antennas coupled to each of the RRUs. The function of the base station may include receiving a UE information, receiving a service flow, deriving a service type of the service flow, and determining a MIMO scheme, modulating the service flow into a MIMO stream according to the MIMO scheme, allocating a plurality of resource blocks, wherein the resource blocks may correspond to a portion or all of the RRUs, distributing the MIMO stream into a plurality of downlink sub streams and transmitting the downlink sub streams through the resource blocks by the RRUs corresponding to the resource blocks.
    Type: Application
    Filed: September 3, 2013
    Publication date: March 13, 2014
    Applicant: Industrial Technology Research Institute
    Inventors: Hsin-An Hou, Wan-Yi Lin, Yung-Han Chen
  • Patent number: 8130378
    Abstract: A phase retardance inspection instrument, comprising: a light source module for generating a single-wavelength light beam; a circularly polarized light generating module, comprising a polarizer and a first phase retarder, for receiving the single-wavelength light beam as it is guided to pass through the polarizer and the first phase retarder in order; and a detecting module, comprising a second phase retarder, a polarizing beam splitter, a first image sensor and a second image sensor, for receiving and guiding a circularly polarized light beam to travel through the second phase retarder and the polarizing beam splitter in order after it passes through a substrate under inspection, wherein the polarizing beam splitter splits an elliptically polarized light beam into intensity vector components of a left-hand circularly polarized light beam and a right-hand circularly polarized light beam, which are to be emitted into the first image sensor and the second image sensor, respectively.
    Type: Grant
    Filed: February 6, 2009
    Date of Patent: March 6, 2012
    Assignee: Industrial Technology Research Institute
    Inventors: Chun-I Wu, Kai-Ping Chuang, Wan-Yi Lin, Yi-Chen Hsieh, Fu-Shiang Yang
  • Publication number: 20110182466
    Abstract: A method and system of detecting colluders conducting a collusion attack including a minority-type collusion attack on a digital product includes the generation of codewords used as watermarks in the digital product. The inner code of the codewords is generated using permutations of rows in a Hadamard matrix and concatenating them together. A typical outer code of the codeword is the Reed Solomon code. An adaptive detector is able to accurately detect one of three or more colluders of a minority-type attack. Prior art schemes using an error correcting code-based watermarking mechanism with an inner code fail to detect colluders with a minority-type collusion attack which includes three colluders.
    Type: Application
    Filed: September 26, 2008
    Publication date: July 28, 2011
    Inventors: Wan-Yi Lin, Shan He, Jeffrey Adam Bloom
  • Publication number: 20110170737
    Abstract: A method and apparatus are described including generating a unique code for each of a plurality of users using a plurality of symbols, generating a plurality of codes representing the plurality of symbols, substituting the plurality of codes into the unique code for each of the plurality of users, permuting the code resulting from the substitution to produce a codeword for each of the plurality of users and embedding the codeword into digital content. The second generating act further includes generating a string of first symbols followed by second symbols, wherein the first symbols are all ones and the second symbols are all negative ones, wherein a number of first symbols is equal to a number of the second symbols, and wherein if a length of the first symbols followed by the second symbols is less than a length of the code, then the first symbols followed by the second symbols are repeated until the code length is filled.
    Type: Application
    Filed: September 26, 2008
    Publication date: July 14, 2011
    Inventors: Wan-yi Lin, Shan He, Jeffrey Adam Bloom