Patents by Inventor Xiaoyu Xiu
Xiaoyu Xiu 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: 20230336785Abstract: An electronic apparatus performs a method of decoding video data. The method includes receiving, from a video bitstream having a hierarchical structure, a first syntax element associated with a first level of the hierarchical structure, in accordance with a determination that the first syntax element indicates that a Cross-Component Sample Adaptive Offset (CCSAO) filter information is present in the first level, reconstructing, from the video bitstream, one or more regions under the first level jointly according to the CCSAO filter information, and in accordance with a determination that the first syntax element indicates that the CCSAO filter information is not present in the first level, reconstructing, from the video bitstream, the one or more regions separately according to the CCSAO filter information present in a second level of the hierarchical structure.Type: ApplicationFiled: June 21, 2023Publication date: October 19, 2023Inventors: Che-Wei Kuo, Xiaoyu Xiu, Wei Chen, Xianglin Wang, Yi-Wen Chen, Hong-Jheng Jhu, Bing Yu
-
Publication number: 20230328293Abstract: A method and an apparatus for image filtering in video coding using a neural network are provided. The method includes: loading a plurality of input patches associated with a current image to be coded, where the plurality of input patches include a first input patch with a first resolution, a second input patch with a second resolution, and a third input patch with a third resolution; and in response to determining that one resolution in the first resolution, the second resolution, and the third resolution is different from the other two resolutions, adjusting the first resolution, the second resolution, and the third resolution at one region of a plurality of regions before the neural network or in the neural network.Type: ApplicationFiled: June 2, 2023Publication date: October 12, 2023Applicant: BEIJING DAJIA INTERNET INFORMATION TECHNOLOGY CO., LTD.Inventors: Wei CHEN, Xiaoyu XIU, Yi-Wen CHEN, Hong-Jheng JHU, Che-Wei KUO, Xianglin WANG, Bing YU
-
Publication number: 20230319315Abstract: An electronic apparatus performs a method of decoding video data. The method includes: receiving, from the video signal, a picture frame that includes a first component and a second component; reconstructing samples of the first component through a first in-loop filter; reconstructing samples of the second component through a second in-loop filter; determining a classifier for the first component from one or more reconstructed samples of the second component relative to a respective reconstructed sample of the first component; selecting a first sample offset for the respective reconstructed sample of the first component according to the classifier; applying a plurality of filters in parallel with the selection of the first sample offset to obtain a plurality of parallel offsets; and clipping an output of a combination of the respective reconstructed sample of the first component, the first sample offset, and the plurality of parallel offsets from the plurality of filters.Type: ApplicationFiled: May 17, 2023Publication date: October 5, 2023Inventors: Che-Wei Kuo, Xiaoyu Xiu, Wei Chen, Xianglin Wang, Yi-Wen Chen, Hong-Jheng Jhu, Ning Yan, Bing Yu
-
Publication number: 20230319314Abstract: An electronic apparatus performs a method of decoding video data, including: reconstructing, from a video bitstream, a picture frame that includes a luma component, a first and a second chroma components, and applying a trained neural network based in-loop filter to the reconstructed picture frame by: converting a first resolution of the samples of the at least one of the first and the second chroma components to a second resolution of the samples of the luma component when the first resolution is different from the second resolution; concatenating samples of at least one of the first and the second chroma components with the luma component; processing the concatenated samples using a convolutional neural network; and reconverting the samples of the at least one of the first and the second chroma components processed by the convolutional neural network from the second resolution back to the first resolution.Type: ApplicationFiled: March 31, 2023Publication date: October 5, 2023Inventors: Tsung-Chuan Ma, Wei Chen, Xiaoyu Xiu, Yi-Wen Chen, Hong-Jheng Jhu, Che-Wei Kuo, Xianglin Wang, Bing Yu
-
Publication number: 20230319289Abstract: Systems and methods are disclosed for video encoding and video decoding using decoder-side intra mode derivation (DIMD). Techniques are provided to code and to decode a video block of a video frame into a bitstream, including determining costs of using respective candidate intra prediction modes to predict samples in a template region adjacent to the video block, deriving an intra prediction mode based on candidates of the candidate intra prediction modes and their respective costs, and predicting samples in the current video block using the derived intra prediction mode. The provided techniques further include determining the costs in multiple stages. In an initial stage, the costs of using respective intra prediction modes from an initial set of candidate modes are determined.Type: ApplicationFiled: June 7, 2023Publication date: October 5, 2023Inventors: Xiaoyu XIU, Yuwen HE, Yan YE
-
Publication number: 20230308652Abstract: An electronic apparatus performs a method of decoding video data. The method includes: receiving, from the video signal, a picture frame that includes a first component and a second component; determining a classifier for the second component from a set of one or more samples of the first component associated with a respective sample of the second component; determining whether to modify a value of the respective sample of the second component of a current block of the picture frame within a virtual boundary according to the classifier; in response to the determination to modify the value of the respective sample of the second component of the current block according to the classifier, determining a sample offset for the respective sample of the second component according to the classifier; and modifying the value of the respective sample of the second component based on the determined sample offset.Type: ApplicationFiled: March 23, 2023Publication date: September 28, 2023Applicant: BEIJING DAJIA INTERNET INFORMATION TECHNOLOGY CO., LTD.Inventors: Che-Wei Kuo, Xiaoyu Xiu, Yi-Wen Chen, Wei Chen, Tsung-Chuan Ma, Hong-Jheng Jhu, Bing Yu, Xianglin Wang
-
Publication number: 20230291936Abstract: Methods, apparatuses, and non-transitory computer-readable storage mediums are provided for video decoding. A decoder may receive a video stream. The decoder may receive a control flag in a slice header level. The decoder may receive at least one syntax element in the slice header level. The decoder may entropy decode the video bitstream based on the control flag and the at least one syntax element.Type: ApplicationFiled: March 23, 2023Publication date: September 14, 2023Applicant: BEIJING DAJIA INTERNET INFORMATION TECHNOLOGY CO., LTD.Inventors: Hong-Jheng JHU, Xiaoyu XIU, Yi-Wen CHEN, Tsung-Chuan MA, Che-Wei KUO, Xianglin WANG, Bing YU
-
Publication number: 20230291906Abstract: Methods, apparatuses, and non-transitory computer-readable storage mediums are provided for video encoding. An encoder may receive a video input. The encoder may obtain a quantization parameter based on the video input. The encoder may derive a rice parameter based on at least one predefined threshold, a coding bit-depth, and the quantization parameter. The encoder may entropy encode a video bitstream based on the rice parameter.Type: ApplicationFiled: May 16, 2023Publication date: September 14, 2023Applicant: BEIJING DAJIA INTERNET INFORMATION TECHNOLOGY CO., LTD.Inventors: Hong-Jheng JHU, Xiaoyu XIU, Yi-Wen CHEN, Wei CHEN, Che-Wei KUO, Xianglin WANG, Bing YU
-
Patent number: 11743493Abstract: A method for controlling bit-width for bi-directional optical flow (BDOF) for video coding includes acquiring, for a current picture, a first reference picture and a second reference picture, the second reference picture being different than the first reference picture, where the first reference picture is displayed before the current picture and the second reference picture is displayed after the current picture. A motion refinement of a coding unit (CU) is calculated by minimizing a difference between a first prediction L0 and a second prediction L1. First gradient values for the first prediction L0 and second gradient values for the second prediction L1 are calculated. A final bi-prediction of the CU is calculated. Also disclosed are an associated apparatus and an associated non-transitory computer readable storage medium.Type: GrantFiled: April 21, 2022Date of Patent: August 29, 2023Assignee: Beijing Dajia Internet Information Technology Co., Ltd.Inventors: Xiaoyu Xiu, Yi-Wen Chen, Xianglin Wang
-
Publication number: 20230262235Abstract: An electronic apparatus performs a method of decoding video data. The method comprises: receiving, from the video signal, a picture frame that includes a first component and a second component; determining a classifier for the second component from a set of samples of the first component associated with a respective sample of the second component; when the set of samples of the first component associated with the respective sample of the second component is divided by a virtual boundary, copying one or more central subsets of the set of samples of the first component to a first boundary position and a second boundary position of the set of samples of the first component; determining a sample offset for the respective sample of the second component according to the classifier; and modifying the value of the respective sample of the second component based on the determined sample offset.Type: ApplicationFiled: April 28, 2023Publication date: August 17, 2023Applicant: BEIJING DAJIA INTERNET INFORMATION TECHNOLOGY CO., LTD.Inventors: Che-Wei KUO, Xiaoyu XIU, Wei CHEN, Xianglin WANG, Yi-Wen CHEN, Hong-Jheng JHU, Bing YU
-
Publication number: 20230254501Abstract: Methods, apparatuses, and non-transitory computer-readable storage mediums are provided for encoding video. An encoder may obtain a first motion vector (MV) associated with a video block obtained from the video. The encoder may obtain a first prediction signal of the video block using the first MV. The encoder may identify a target MV by applying a gradient-based motion refinement algorithm in a recursive manner using the first prediction signal and the first MV. The encoder may obtain a second prediction signal of the video block based on the target MV.Type: ApplicationFiled: April 14, 2023Publication date: August 10, 2023Applicant: BEIJING DAJIA INTERNET INFORMATION TECHNOLOGY CO., LTD.Inventors: Xiaoyu XIU, Che-Wei KUO, Wei CHEN, Yi-Wen CHEN, Hong-Jheng JHU, Xianglin WANG, Bing YU
-
Publication number: 20230252300Abstract: A method and an apparatus for hybrid training of neural networks for video coding are provided. The method includes: obtaining, in an offline training stage, an offline trained network by training a first neural network offline using a plurality of first data sets; refining, in an online training stage, a plurality of neural network layers using a plurality of second data sets, wherein the plurality of neural network layers comprise at least one neural network layer in the offline trained network or in a second neural network connected to the offline trained network.Type: ApplicationFiled: March 8, 2023Publication date: August 10, 2023Applicant: BEIJING DAJIA INTERNET INFORMATION TECHNOLOGY CO., LTD.Inventors: Wei CHEN, Xiaoyu Xiu, Yi-Wen Chen, Hong-Jheng Jhu, Che-Wei Kuo, Xianglin Wang, Bing Yu
-
Patent number: 11722677Abstract: Systems and methods related to video encoding and decoding using decoder-side intra mode derivation (DIMD) are described. In an exemplary method of coding samples in a block in a video, an intra coding mode is selected based on a plurality of reconstructed samples in a template region adjacent to the block, and the samples in the block are predicted with intra prediction using the selected intra coding mode. The intra coding mode may be selected by testing a plurality of candidate intra coding modes for cost (e.g. distortion) of predicting the template region from a set of reconstructed reference samples. The mode with the lowest cost is used for prediction. In exemplary embodiments, explicit signaling of the intra mode is not required.Type: GrantFiled: May 5, 2017Date of Patent: August 8, 2023Assignee: InterDigital Madison Patent Holdings, SASInventors: Xiaoyu Xiu, Yuwen He, Yan Ye
-
Publication number: 20230209093Abstract: An electronic apparatus performs a method of decoding video data. The method comprises: receiving, from the video signal, a picture frame that includes a first component and a second component; receiving, from the video signal, a first syntax element that indicates whether Cross-component Sample Adaptive Offset (CCSAO) is enabled for the second component at a picture level; when the first syntax element indicates the CCSAO is enabled for the second component at the picture level, receiving, from the video signal, a second syntax element that indicates whether the CCSAO is controlled for the second component at a coding tree block (CTB) level; and when the second syntax element indicates the CCSAO is controlled for the second component at the CTB level, receiving, from the video signal, a third syntax element that indicates whether the CCSAO is enabled for the second component at the CTB level.Type: ApplicationFiled: February 21, 2023Publication date: June 29, 2023Inventors: Che-Wei Kuo, Xiaoyu Xiu, Wei Chen, Xianglin Wang, Yi-Wen Chen, Tsung-Chuan Ma, Hong-Jheng Jhu, Bing Yu
-
Publication number: 20230199223Abstract: A method, apparatus, and a non-transitory computer-readable storage medium for decoding a video signal are provided. A decoder may receive, through a bitstream, arranged syntax elements in sequence parameter set (SPS) level. The arranged syntax elements in the SPS level are arranged so that functions of related syntax elements are grouped in versatile video coding (VVC) syntax at a coding level. The decoder may receive, through the bitstream and in response to multiple syntax elements satisfy a predefined condition, a second syntax element immediately after the multiple syntax elements. The decoder may perform, through the bitstream, a related syntax element function to video data from the bitstream in accordance with the multiple syntax elements and the second syntax element.Type: ApplicationFiled: October 31, 2022Publication date: June 22, 2023Applicant: BEIJING DAJIA INTERNET INFORMATION TECHNOLOGY CO., LTD.Inventors: Hong-Jheng JHU, Yi-Wen CHEN, Xiaoyu XIU, Tsung-Chuan MA, Wei CHEN, Xianglin WANG, Bing YU
-
Publication number: 20230199209Abstract: An electronic apparatus performs a method of decoding video data. The method includes receiving, from the video signal, a picture frame that includes a first component and a second component, receiving, from the video signal, a plurality of sample offsets associated with the second component, reconstructing the samples of the first component before a first in-loop filter module, reconstructing the samples of the second component after a second in-loop filter module, determining a classifier for the second component from one or more reconstructed samples of the first component relative to each sample of the second component, selecting a sample offset from the plurality of sample offsets for the second component according to the classifier, and modifying the reconstructed samples of the second component based on the selected sample offset.Type: ApplicationFiled: February 23, 2023Publication date: June 22, 2023Inventors: Che-Wei Kuo, Xiaoyu Xiu, Wei Chen, Xianglin Wang, Yi-Wen Chen, Tsung-Chuan Ma, Hong-Jheng Jhu, Bing Yu
-
Publication number: 20230199219Abstract: A system, method, and/or instrumentality may be provided for coding a 360-degree video. A picture of the 360-degree video may be received. The picture may include one or more faces associated with one or more projection formats. A first projection format indication may be received that indicates a first projection format may be associated with a first face. A second projection format indication may be received that indicates a second projection format may be associated with a second face. Based on the first projection format, a first transform function associated with the first face may be determined. Based on the second projection format, a second transform function associated with the second face may be determined. At least one decoding process may be performed on the first face using the first transform function and/or at least one decoding process may be performed on the second face using the second transform function.Type: ApplicationFiled: February 21, 2023Publication date: June 22, 2023Applicant: VID SCALE, INC.Inventors: Xiaoyu Xiu, Yuwen He, Yan Ye
-
Publication number: 20230188748Abstract: A video coding device may be configured to perform directional Bi-directional optical flow (BDOF) refinement on a coding unit (CU). The device may determine the direction in which to perform directional BDOF refinement. The device may calculate the vertical direction gradient difference and the horizontal direction gradient difference for the CU. The vertical direction gradient difference may indicate the difference between the vertical gradients for a first reference picture and the vertical gradients for a second reference picture. The horizontal direction gradient difference may indicate the difference between the horizontal gradients for the first reference picture and the horizontal gradients for the second reference picture. The video coding device may determine the direction in which to perform directional BDOF refinement based on the vertical direction gradient difference and the horizontal direction gradient difference.Type: ApplicationFiled: February 6, 2023Publication date: June 15, 2023Applicant: VID SCALE, INC.Inventors: Yuwen He, Xiaoyu Xiu, Yan Ye
-
Publication number: 20230179795Abstract: Systems, methods, and instrumentalities may be provided for determining whether to bypass bi-directional optical flow (BDOF) if BDOF is used in combination with bi-prediction with coding unit (CU) weights (e.g., generalized bi-prediction (GBi)). A coding system may combine coding modes, coding techniques, and/or coding tools. The coding system may include a wireless transmit/receive unit (WTRU). For example, the coding system may combine BDOF and bi-prediction with CU weights (BCW). BDOF may include refining a motion vector associated with a current CU based at least in part on gradients associated with a location in the current CU. The coding system may determine that BDOF is enabled, and/or that bi-prediction with CU weights is enabled for the current CU. The coding system’s determination that bi-prediction with CU weights is enabled and/or that BDOF is enabled may be based on one or more indications.Type: ApplicationFiled: January 9, 2023Publication date: June 8, 2023Applicant: VID SCALE, INC.Inventors: Yuwen He, Jiancong Luo, Xiaoyu Xiu, Yan Ye
-
Publication number: 20230156215Abstract: Systems, methods, and instrumentalities for sub-block motion derivation and motion vector refinement for merge mode may be disclosed herein. Video data may be coded (e.g., encoded and/or decoded). A collocated picture for a current slice of the video data may be identified. The current slice may include one or more coding units (CUs). One or more neighboring CUs may be identified for a current CU. A neighboring CU (e.g., each neighboring CU) may correspond to a reference picture. A (e.g., one) neighboring CU may be selected to be a candidate neighboring CU based on the reference pictures and the collocated picture. A motion vector (MV) (e.g., collocated MV) may be identified from the collocated picture based on an MV (e.g., a reference MV) of the candidate neighboring CU. The current CU may be coded (e.g., encoded and/or decoded) using the collocated MV.Type: ApplicationFiled: January 16, 2023Publication date: May 18, 2023Applicant: VID SCALE, INC.Inventors: Xiaoyu Xiu, Yuwen He, Yan Ye