公开(公告)号：US20180192073A1

公开(公告)日：2018-07-05

申请号：US15907026

申请日：2018-02-27

Applicant: HFI Innovation Inc.

Inventor： Mei Guo ,  Shan Liu ,  Shaw-Min Lei

IPC: H04N19/593 ,  H04N19/122 ,  H04N19/61 ,  H04N19/33 ,  H04N19/176 ,  H04N19/11

CPC classification number: H04N19/593 ,  H04N19/11 ,  H04N19/122 ,  H04N19/176 ,  H04N19/33 ,  H04N19/61

Abstract: A method and apparatus of predictive Intra coding of a block are disclosed. In one embodiment according to the present invention, a set of Intra prediction modes is determined, and predictive Intra coding is applied to a prediction unit (PU) of an image based on one or more neighboring PUs according to a current Intra prediction mode. The current Intra prediction mode is selected from the set of Intra prediction modes and the current Intra prediction mode is predicted using most probable modes (MPMs) and remaining modes. A number of the remaining modes is kept as a power of 2. For example, the 32 remaining modes can be represented using 5-bit fixed length codes.

公开(公告)号：US10009612B2

公开(公告)日：2018-06-26

申请号：US14378969

申请日：2013-04-12

Applicant: HFI INNOVATION INC.

Inventor： Shan Liu ,  Shaw-Min Lei

IPC: H04N19/159 ,  H04N19/96 ,  H04N19/186 ,  H04N19/61 ,  H04N19/122 ,  H04N19/176 ,  H04N19/157

CPC classification number: H04N19/159 ,  H04N19/122 ,  H04N19/157 ,  H04N19/176 ,  H04N19/186 ,  H04N19/61 ,  H04N19/96

Abstract: A method and apparatus for video data processing for video in YUV422 or YUV 444 formats are disclosed. In one embodiment, for a 2N×2N luma coding unit (CU) in YUV422 format, the transform process partitions residue data corresponding to the 2N×2N luma CU and the N×2N chroma CU into square luma and chroma transform units (TUs). The residue data associated with the luma and the chroma CUs are generated by applying prediction process to the luma CU and the chroma CU. The transform process is independent of prediction block size or prediction mode associated with the prediction process. In another embodiment, the prediction process splits the CU into two prediction blocks. Transform process is applied on the chroma residue data corresponding to the chroma CU to form one or more chroma TUs, wherein the transform process is dependent on CU size and prediction block size, or CU size and prediction mode.

公开(公告)号：US09681135B2

公开(公告)日：2017-06-13

申请号：US15123866

申请日：2015-03-13

Applicant: HFI Innovation Inc.

Inventor： Wang-Lin Lai ,  Yu-Chen Sun ,  Shan Liu ,  Xiaozhong Xu ,  Tzu-Der Chuang ,  Ding-Yun Chen ,  Yu-Wen Huang

IPC: H04N19/11 ,  H04N1/64 ,  H04N19/176 ,  H04N19/593 ,  H04N19/463 ,  H04N19/186 ,  H04N19/154 ,  H04N19/94

CPC classification number: H04N19/11 ,  H04N1/644 ,  H04N19/105 ,  H04N19/154 ,  H04N19/159 ,  H04N19/176 ,  H04N19/186 ,  H04N19/197 ,  H04N19/426 ,  H04N19/463 ,  H04N19/593 ,  H04N19/94

Abstract: A method and apparatus for palette coding of a block of video data using a shared palette/palette prediction buffer are disclosed. The system receives a bitstream including compressed data associated with a current palette coded block and determines a palette predictor for the current palette coded block stored in a shared palette/palette predictor buffer. Reuse flags for entries of the current palette are used to update the current palette stored in the shared palette/palette predictor buffer. Any new signaled entry will be inserted to the shared palette/palette predictor buffer. An updated palette predictor for a next palette coded block is generated accordingly. A method of palette management and initialization for first palette-coded block in a high-level video structure is also disclosed. The high-level video structure may correspond to a slice, wavefront, sequence or picture.

公开(公告)号：US09596479B2

公开(公告)日：2017-03-14

申请号：US14821100

申请日：2015-08-07

Applicant: HFI Innovation Inc.

Inventor： Shan Liu ,  Jungsun Kim ,  Tzu-Der Chuang

IPC: G06K9/00 ,  H04N19/50 ,  H04N19/96 ,  H04N19/46 ,  H04N1/64 ,  H04N19/105 ,  H04N19/186 ,  H04N19/70 ,  H04N19/176 ,  H04N19/593 ,  H04N19/11 ,  H04N19/463 ,  H04N19/182 ,  H04N19/93 ,  H04N19/94

CPC classification number: H04N19/50 ,  H04N1/644 ,  H04N19/105 ,  H04N19/11 ,  H04N19/176 ,  H04N19/182 ,  H04N19/186 ,  H04N19/46 ,  H04N19/463 ,  H04N19/593 ,  H04N19/70 ,  H04N19/93 ,  H04N19/94 ,  H04N19/96

Abstract: Methods for coding a block of video data using palette coding in a video coding are disclosed. According to one embodiment, the palette transpose flag is signaled conditionally depending on the maximum index value or the palette size. If the maximum index value or the palette size is equal to or greater than a threshold size, the palette transpose flag is signaled at an encoder side or parsed at a decoder side. Otherwise, the palette transpose flag is not signaled or parsed. In another embodiment, the last palette and the last palette size are updated by the current palette with the current palette size conditionally. For example, updating the last palette and the last palette size by the current palette with the current palette size is skipped if the current coding unit is coded using a pulse-coded modulation (PCM) mode or the current palette size is zero.

Abstract translation: 公开了使用视频编码中的调色板编码对视频数据块进行编码的方法。 根据一个实施例，根据最大索引值或调色板大小有条件地发信号通知调色板转置标志。 如果最大索引值或调色板大小等于或大于阈值大小，则调色板转置标志在编码器侧发信号通知或在解码器侧进行解析。 否则，调色板转置标志不发信号或解析。 在另一个实施例中，最后一个调色板和最后一个调色板大小由具有条件的当前调色板大小的当前调色板更新。 例如，如果使用脉冲编码调制（PCM）模式或当前调色板大小为零对当前编码单元进行编码，则跳过具有当前调色板大小的当前调色板更新最后调色板和最后调色板尺寸。

公开(公告)号：US11297352B2

公开(公告)日：2022-04-05

申请号：US16930088

申请日：2020-07-15

Applicant: HFI Innovation Inc.

Inventor： Xiaozhong Xu ,  Shan Liu

IPC: H04N19/86 ,  H04N19/159 ,  H04N19/176 ,  H04N19/117 ,  H04N19/182 ,  H04N19/105 ,  H04N19/196

Abstract: A method and apparatus are directed to applying deblocking filter to block boundaries in a video coding system including an IntraBC (Intra-block copy) mode. For at least one combination of two neighboring reconstructed blocks including at least one IntraBC coded block in the two neighboring reconstructed blocks, a boundary filter strength equal to one or zero is selected based on coding parameters associated with the two neighboring reconstructed blocks. Deblocking filter is then applied to neighboring samples of the two neighboring reconstructed blocks around the block boundary using the boundary filter strength selected. For example, when the two neighboring reconstructed blocks correspond to one IntraBC coded block and one Inter coded block, the boundary filter strength is set to one.

公开(公告)号：US10715801B2

公开(公告)日：2020-07-14

申请号：US15598454

申请日：2017-05-18

Applicant: HFI Innovation Inc.

Inventor： Wang-Lin Lai ,  Yu-Chen Sun ,  Shan Liu ,  Xiaozhong Xu ,  Tzu-Der Chuang ,  Ding-Yun Chen ,  Yu-Wen Huang

IPC: H04N19/11 ,  H04N19/176 ,  H04N19/463 ,  H04N19/105 ,  H04N19/196 ,  H04N19/154 ,  H04N19/593 ,  H04N19/186 ,  H04N19/159 ,  H04N19/426 ,  H04N19/94 ,  H04N1/64

Abstract: A method of palette management for palette coding in a video coding system receives input data associated with a current block in a high-level picture structure and initializes a palette predictor in the high-level picture structure before a corresponding palette of a first palette-coded block in the high-level picture structure is coded. If a palette mode is selected for the current block, the method applies the palette coding to the current block using a current palette and updates the palette predictor based on the current palette to generate an updated palette predictor for a next block coded in the palette mode.

公开(公告)号：US10554979B2

公开(公告)日：2020-02-04

申请号：US15389743

申请日：2016-12-23

Applicant: HFI Innovation Inc.

Inventor： Tzu-Der Chuang ,  Yu-Chen Sun ,  Jungsun Kim ,  Shan Liu ,  Wang-Lin Lai ,  Yi-Wen Chen

IPC: H04N19/157 ,  H04N19/50 ,  H04N19/11 ,  H04N19/182 ,  H04N19/105 ,  H04N19/184 ,  H04N19/176

Abstract: In a palette prediction that includes a copy-above mode for predicting one or more first pixels based on one or more above pixels located above said one or more first pixels correspondingly, a method of palette mode coding using palette prediction for a picture includes receiving input data associated with a current block and determining a current coding mode for one or more current pixels. If the current coding mode corresponds to the copy-above mode, then the method identifies one or more above pixels located above said one or more current pixels correspondingly, wherein said one or more above pixels are restricted to be non-Escape pixels. The method further encodes the one or more current pixels by using said one or more above pixels as predictors.

公开(公告)号：US10455231B2

公开(公告)日：2019-10-22

申请号：US15514129

申请日：2015-09-30

Applicant: HFI Innovation Inc.

Inventor： Xiaozhong Xu ,  Kai Zhang ,  Shan Liu ,  Jicheng An ,  Xianguo Zhang

IPC: H04N19/117 ,  H04N19/52 ,  H04N19/139 ,  H04N19/176 ,  H04N19/124 ,  H04N19/523 ,  H04N19/86

Abstract: A method of MVP (motion vector prediction) for video coding with adaptive motion vector resolution is disclosed. According to the present invention, the MVP coding is applied to the current MV or the current MV is stored depending on the current MV resolution, the reference MV resolution, or both the current MV resolution and the reference MV resolution. In one embodiment, when the current MV resolution corresponds to integer pixel resolution, MVP coding is then applied to the current MV using a modified temporal MV as a motion vector predictor for the current MV, where the modified temporal MV is generated by right-shifting the temporal MV. In another embodiment, when the current MV resolution corresponds to integer pixel resolution, the current MV is left-shifted before it is stored in a memory.

公开(公告)号：US10230981B2

公开(公告)日：2019-03-12

申请号：US15150545

申请日：2016-05-10

Applicant: HFI INNOVATION INC.

Inventor： Ximin Zhang ,  Shan Liu

IPC: H04N19/11 ,  H04N19/593 ,  H04N19/50 ,  H04N19/61 ,  H04N19/463

Abstract: A method is performed by a CODEC (coder/decoder) for Intra prediction mode processing based on multiple MPMs (most probable modes) in video coding. The method includes operations of determining a first Intra mode associated with a left neighboring block of a current block; determining a second Intra mode associated with a top neighboring block of the current block; deriving three MPMs based on the first Intra mode and the second Intra mode, and processing a current Intra mode associated with the current block according to said three MPMs. The step of deriving said three MPMs based on the first Intra mode and the second Intra mode includes: determining whether the first Intra mode is identical to the second Intra mode; and if the first Intra mode is identical to the second Intra mode, deriving said three MPMs according to whether the first Intra mode is identical to a predetermined mode.

公开(公告)号：US10218977B2

公开(公告)日：2019-02-26

申请号：US15384864

申请日：2016-12-20

Applicant: HFI Innovation Inc.

Inventor： Ximin Zhang ,  Shan Liu

IPC: H04B1/66 ,  H04N19/122 ,  G06K9/36 ,  H04N19/124 ,  H04N19/115 ,  H04N11/02 ,  G06K9/32 ,  H04N19/126 ,  H04N19/196 ,  H04N19/176 ,  H04N19/70 ,  H04N19/60 ,  H04N19/18 ,  H04N19/80

Abstract: A method and system process transform blocks according to quantization matrices in a video coding system. The method includes obtaining an initial quantization matrix having a first width and a first height, obtaining a derived quantization matrix having a second width and a second height, wherein the second width is different from the second height, and the derived quantization matrix is derived from the initial quantization matrix, receiving a transform block having a block size, where the transform block is associated with a picture, and selecting the initial quantization matrix or the derived quantization matrix for processing transform coefficients of the transform block according to the block size.