Patents by Inventor Akira Fujibayashi

Akira Fujibayashi 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: 10298953
    Abstract: A moving image encoding/decoding system may include a video predictive encoding device, which may include: an encoding device which encodes each of a plurality of input pictures to generate compressed picture data including a random access picture, and encodes data about display order information of each picture; a restoration device which decodes the compressed picture data to restore a reproduced picture; a picture storage device which stores the reproduced picture as a reference picture; and a memory management device which controls the picture storage device. Following completion of an encoding process of generating the random access picture, the memory management device refreshes the picture storage device by setting every reference picture in the picture storage device, except for the random access picture, as unnecessary immediately before or immediately after encoding a picture with display order information larger than the display order information of the random access picture.
    Type: Grant
    Filed: December 7, 2018
    Date of Patent: May 21, 2019
    Assignee: NTT DOCOMO, INC
    Inventors: Choong Seng Boon, Yoshinori Suzuki, Akira Fujibayashi, Thiow Keng Tan
  • Publication number: 20190116378
    Abstract: A moving image encoding/decoding system may include a video predictive encoding device, which may include: an encoding device which encodes each of a plurality of input pictures to generate compressed picture data including a random access picture, and encodes data about display order information of each picture; a restoration device which decodes the compressed picture data to restore a reproduced picture; a picture storage device which stores the reproduced picture as a reference picture; and a memory management device which controls the picture storage device. Following completion of an encoding process of generating the random access picture, the memory management device refreshes the picture storage device by setting every reference picture in the picture storage device, except for the random access picture, as unnecessary immediately before or immediately after encoding a picture with display order information larger than the display order information of the random access picture.
    Type: Application
    Filed: December 7, 2018
    Publication date: April 18, 2019
    Applicant: NTT DOCOMO, INC.
    Inventors: Choong Seng Boon, Yoshinori Suzuki, Akira Fujibayashi, Thiow Keng Tan
  • Publication number: 20190110069
    Abstract: A moving image encoding/decoding system may include a video predictive encoding device, which may include: an encoding device which encodes each of a plurality of input pictures to generate compressed picture data including a random access picture, and encodes data about display order information of each picture; a restoration device which decodes the compressed picture data to restore a reproduced picture; a picture storage device which stores the reproduced picture as a reference picture; and a memory management device which controls the picture storage device. Following completion of an encoding process of generating the random access picture, the memory management device refreshes the picture storage device by setting every reference picture in the picture storage device, except for the random access picture, as unnecessary immediately before or immediately after encoding a picture with display order information larger than the display order information of the random access picture.
    Type: Application
    Filed: December 7, 2018
    Publication date: April 11, 2019
    Applicant: NTT DOCOMO, INC.
    Inventors: Choong Seng Boon, Yoshinori Suzuki, Akira Fujibayashi, Thiow Keng Tan
  • Publication number: 20190110065
    Abstract: A predictive coding system can encode motion vectors by determining a motion vector of a target partition in a frame picture of an encoding target; determining a first motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to a left-neighboring region to the target partition; determining a second motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to an above-neighboring region to the target partition; selecting an optimum motion vector predictor and outputting motion vector predictor indication information to specify the selected optimum motion vector predictor, the optimum motion vector predictor selected based on comparison between one or more motion vector predictor candidates including the first motion vector predictor candidate and the second motion vector predictor candidate, and the motion vector of the target partition; and encoding the motion vector predictor indication information.
    Type: Application
    Filed: November 7, 2018
    Publication date: April 11, 2019
    Applicant: NTT DOCOMO, INC.
    Inventors: Akira Fujibayashi, Yoshinori Suzuki, Choong Seng Boon
  • Publication number: 20190075316
    Abstract: A predictive coding system can encode motion vectors by determining a motion vector of a target partition in a frame picture of an encoding target; determining a first motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to a left-neighboring region to the target partition; determining a second motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to an above-neighboring region to the target partition; selecting an optimum motion vector predictor and outputting motion vector predictor indication information to specify the selected optimum motion vector predictor, the optimum motion vector predictor selected based on comparison between one or more motion vector predictor candidates including the first motion vector predictor candidate and the second motion vector predictor candidate, and the motion vector of the target partition; and encoding the motion vector predictor indication information.
    Type: Application
    Filed: November 7, 2018
    Publication date: March 7, 2019
    Applicant: NTT DOCOMO, INC.
    Inventors: Akira Fujibayashi, Yoshinori Suzuki, Choong Seng Boon
  • Publication number: 20190028730
    Abstract: A video prediction encoding device provided with a picture storage unit storing, as reference pictures to be used for encoding of a subsequent picture, one or more of reconstructed pictures restored from compressed picture data and a memory control unit to control a frame memory in the picture storage unit. When a frame size of input pictures is not more than 1/(2?L) (where L is an integer of one or more) of a maximum frame size of pictures, the memory control unit determines that (2?L) times the maximum number of reconstructed pictures are storable in the frame memory.
    Type: Application
    Filed: September 26, 2018
    Publication date: January 24, 2019
    Inventors: Akira FUJIBAYASHI, Choong Seng BOON, Thiow Keng TAN
  • Patent number: 10178407
    Abstract: A moving image encoding/decoding system may include a video predictive encoding device, which may include: an encoding device which encodes each of a plurality of input pictures to generate compressed picture data including a random access picture, and encodes data about display order information of each picture; a restoration device which decodes the compressed picture data to restore a reproduced picture; a picture storage device which stores the reproduced picture as a reference picture; and a memory management device which controls the picture storage device. Following completion of an encoding process of generating the random access picture, the memory management device refreshes the picture storage device by setting every reference picture in the picture storage device, except for the random access picture, as unnecessary immediately before or immediately after encoding a picture with display order information larger than the display order information of the random access picture.
    Type: Grant
    Filed: August 10, 2015
    Date of Patent: January 8, 2019
    Assignee: NTT DOCOMO, INC
    Inventors: Choong Seng Boon, Yoshinori Suzuki, Akira Fujibayashi, Thiow Keng Tan
  • Publication number: 20190007694
    Abstract: A video predictive encoding device can include an input device to implement input of pictures constituting a video sequence; an encoder to encode the pictures by intra-prediction or inter-prediction and to encode output order information of each picture and information about a picture type of each picture to generate compressed picture data. The compressed picture data includes a picture to serve as a random access point. The video predictive coding device can also include a reconstruction unit to decode the compressed picture data to reconstruct pictures; a picture storage unit to store one or more of the reconstructed pictures as reference pictures to be used for encoding of a subsequent picture; and a controller to determine the picture type and control the picture storage unit based on the determined picture type. The controller can label each picture as one of following three types: CRA picture, type-1 picture and type-2 picture.
    Type: Application
    Filed: August 20, 2018
    Publication date: January 3, 2019
    Inventors: Choong Seng Boon, Junya Takiue, Akira Fujibayashi, Thiow Keng Tan
  • Patent number: 10154280
    Abstract: A predictive coding system can encode motion vectors by determining a motion vector of a target partition in a frame picture of an encoding target; determining a first motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to a left-neighboring region to the target partition; determining a second motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to an above-neighboring region to the target partition; selecting an optimum motion vector predictor and outputting motion vector predictor indication information to specify the selected optimum motion vector predictor, the optimum motion vector predictor selected based on comparison between one or more motion vector predictor candidates including the first motion vector predictor candidate and the second motion vector predictor candidate, and the motion vector of the target partition; and encoding the motion vector predictor indication information.
    Type: Grant
    Filed: January 16, 2018
    Date of Patent: December 11, 2018
    Assignee: NTT DOCOMO, INC.
    Inventors: Akira Fujibayashi, Yoshinori Suzuki, Choong Seng Boon
  • Patent number: 10116956
    Abstract: A video prediction encoding device provided with a picture storage unit storing, as reference pictures to be used for encoding of a subsequent picture, one or more of reconstructed pictures restored from compressed picture data and a memory control unit to control a frame memory in the picture storage unit. When a frame size of input pictures is not more than 1/(2^L) (where L is an integer of one or more) of a maximum frame size of pictures, the memory control unit determines that (2^L) times the maximum number of reconstructed pictures are storable in the frame memory.
    Type: Grant
    Filed: October 25, 2017
    Date of Patent: October 30, 2018
    Assignee: NTT DOCOMO, INC.
    Inventors: Akira Fujibayashi, Choong Seng Boon, Thiow Keng Tan
  • Patent number: 10075724
    Abstract: A video predictive encoding device can include an input device to implement input of pictures constituting a video sequence; an encoder to encode the pictures by intra-prediction or inter-prediction and to encode output order information of each picture and information about a picture type of each picture to generate compressed picture data. The compressed picture data includes a picture to serve as a random access point. The video predictive coding device can also include a reconstruction unit to decode the compressed picture data to reconstruct pictures; a picture storage unit to store one or more of the reconstructed pictures as reference pictures to be used for encoding of a subsequent picture; and a controller to determine the picture type and control the picture storage unit based on the determined picture type. The controller can label each picture as one of following three types: CRA picture, type-1 picture and type-2 picture.
    Type: Grant
    Filed: December 26, 2014
    Date of Patent: September 11, 2018
    Assignee: NTT DOCOMO, INC.
    Inventors: Choong Seng Boon, Junya Takiue, Akira Fujibayashi, Thiow Keng Tan
  • Publication number: 20180146209
    Abstract: A predictive coding system can encode motion vectors by determining a motion vector of a target partition in a frame picture of an encoding target; determining a first motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to a left-neighboring region to the target partition; determining a second motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to an above-neighboring region to the target partition; selecting an optimum motion vector predictor and outputting motion vector predictor indication information to specify the selected optimum motion vector predictor, the optimum motion vector predictor selected based on comparison between one or more motion vector predictor candidates including the first motion vector predictor candidate and the second motion vector predictor candidate, and the motion vector of the target partition; and encoding the motion vector predictor indication information.
    Type: Application
    Filed: January 16, 2018
    Publication date: May 24, 2018
    Applicant: NTT DOCOMO, Inc.
    Inventors: Akira Fujibayashi, Yoshinori Suzuki, Choong Seng Boon
  • Patent number: 9942562
    Abstract: A video prediction encoding device provided with a picture storage unit storing, as reference pictures to be used for encoding of a subsequent picture, one or more of reconstructed pictures restored from compressed picture data and a memory control unit to control a frame memory in the picture storage unit. When a frame size of input pictures is not more than 1/(2^L) (where L is an integer of one or more) of a maximum frame size of pictures, the memory control unit determines that (2^L) times the maximum number of reconstructed pictures are storable in the frame memory.
    Type: Grant
    Filed: December 29, 2014
    Date of Patent: April 10, 2018
    Assignee: NTT DOCOMO, INC.
    Inventors: Akira Fujibayashi, Choong Seng Boon, Thiow Keng Tan
  • Publication number: 20180048906
    Abstract: A video prediction encoding device provided with a picture storage unit storing, as reference pictures to be used for encoding of a subsequent picture, one or more of reconstructed pictures restored from compressed picture data and a memory control unit to control a frame memory in the picture storage unit. When a frame size of input pictures is not more than 1/(2?L) (where L is an integer of one or more) of a maximum frame size of pictures, the memory control unit determines that (2?L) times the maximum number of reconstructed pictures are storable in the frame memory.
    Type: Application
    Filed: October 25, 2017
    Publication date: February 15, 2018
    Inventors: Akira FUJIBAYASHI, Choong Seng BOON, Thiow Keng TAN
  • Patent number: 9877040
    Abstract: A predictive coding system can encode motion vectors by determining a motion vector of a target partition in a frame picture of an encoding target; determining a first motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to a left-neighboring region to the target partition; determining a second motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to an above-neighboring region to the target partition; selecting an optimum motion vector predictor and outputting motion vector predictor indication information to specify the selected optimum motion vector predictor, the optimum motion vector predictor selected based on comparison between one or more motion vector predictor candidates including the first motion vector predictor candidate and the second motion vector predictor candidate, and the motion vector of the target partition; and encoding the motion vector predictor indication information.
    Type: Grant
    Filed: December 9, 2016
    Date of Patent: January 23, 2018
    Assignee: NTT DOCOMO, INC.
    Inventors: Akira Fujibayashi, Yoshinori Suzuki, Choong Seng Boon
  • Patent number: 9693052
    Abstract: A predictive coding system can encode motion vectors by determining a motion vector of a target partition in a frame picture of an encoding target; determining a first motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to a left-neighboring region to the target partition; determining a second motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to an above-neighboring region to the target partition; selecting an optimum motion vector predictor and outputting motion vector predictor indication information to specify the selected optimum motion vector predictor, the optimum motion vector predictor selected based on comparison between one or more motion vector predictor candidates including the first motion vector predictor candidate and the second motion vector predictor candidate, and the motion vector of the target partition; and encoding the motion vector predictor indication information.
    Type: Grant
    Filed: July 5, 2013
    Date of Patent: June 27, 2017
    Assignee: NTT DoCoMo, Inc.
    Inventors: Akira Fujibayashi, Yoshinori Suzuki, Choong Seng Boon
  • Publication number: 20170155898
    Abstract: A predictive coding system can encode motion vectors by determining a motion vector of a target partition in a frame picture of an encoding target; determining a first motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to a left-neighboring region to the target partition; determining a second motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to an above-neighboring region to the target partition; selecting an optimum motion vector predictor and outputting motion vector predictor indication information to specify the selected optimum motion vector predictor, the optimum motion vector predictor selected based on comparison between one or more motion vector predictor candidates including the first motion vector predictor candidate and the second motion vector predictor candidate, and the motion vector of the target partition; and encoding the motion vector predictor indication information.
    Type: Application
    Filed: July 5, 2013
    Publication date: June 1, 2017
    Inventors: Akira Fujibayashi, Yoshinori Suzuki, Choong Seng Boon
  • Publication number: 20170150169
    Abstract: A predictive coding system can encode motion vectors by determining a motion vector of a target partition in a frame picture of an encoding target; determining a first motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to a left-neighboring region to the target partition; determining a second motion vector predictor candidate from a motion vector or motion vectors of one or more partitions belonging to an above-neighboring region to the target partition; selecting an optimum motion vector predictor and outputting motion vector predictor indication information to specify the selected optimum motion vector predictor, the optimum motion vector predictor selected based on comparison between one or more motion vector predictor candidates including the first motion vector predictor candidate and the second motion vector predictor candidate, and the motion vector of the target partition; and encoding the motion vector predictor indication information.
    Type: Application
    Filed: December 9, 2016
    Publication date: May 25, 2017
    Inventors: Akira Fujibayashi, Yoshinori Suzuki, Choong Seng Boon
  • Patent number: 9654415
    Abstract: This invention includes an application server for executing an application and transmitting the execution results to a terminal or other such device connected to a network, and a management server for allocating the application and data on an application server and a storage device, respectively, wherein the management server reallocates the application execution site using: a procedure for obtaining device location information from a device via an application server and selecting, from the location information, an application server that will be the application migration destination; a procedure for indicating the migration-destination application server to the migration-source application server; and a procedure for migrating the application and the data between the application servers.
    Type: Grant
    Filed: April 20, 2012
    Date of Patent: May 16, 2017
    Assignee: HITACHI, LTD.
    Inventors: Takatoshi Kato, Yutaka Kudo, Naoto Matsunami, Akira Fujibayashi
  • Patent number: 9513825
    Abstract: Storage system: wherein processor number information includes at least one logical unit number and at least one processor number of storage nodes; wherein transfer list index/processor number information includes a processor number for identifying a processor from among processors of the plurality of storage nodes, and index information for identifying a transfer list including instruction which the processor sends to the protocol processor; wherein a local router determines a first processor from among the processors of the plurality of storage nodes which is to be a transfer destination of a write request based on processor number information in response to the write request from the host computer through the protocol processor; wherein the first processor generates and sends to the protocol processor a first transfer list which includes instruction for processing, and generates first index information which is an index of the first transfer list upon receiving the write request.
    Type: Grant
    Filed: September 22, 2014
    Date of Patent: December 6, 2016
    Assignee: HITACHI, LTD.
    Inventors: Shuji Nakamura, Akira Fujibayashi, Mutsumi Hosoya