公开(公告)号：US20190068994A1

公开(公告)日：2019-02-28

申请号：US15824058

申请日：2017-11-28

Applicant: GOOGLE LLC

Inventor： Dake He

IPC: H04N19/91 ,  H04N19/96 ,  H04N19/176 ,  H04N19/124 ,  H04N19/13

CPC classification number: H04N19/13 ,  H04N19/124 ,  H04N19/129 ,  H04N19/159 ,  H04N19/176 ,  H04N19/18 ,  H04N19/196 ,  H04N19/91 ,  H04N19/96

Abstract: Entropy encoding and decoding a sequence of symbols using probability mixing is disclosed. A method includes selecting models that include a first model and a second model, for at least a symbol, at a position of the symbols, determining a mixed probability using the first model and the second model, and coding the symbol using the mixed probability. Determining the mixed probability for the symbol includes determining, using the first model, a first conditional probability for coding the symbol, determining, using the second model, a second conditional probability for coding the symbol, and determining, using the first conditional probability and the second conditional probability, the mixed probability for coding the symbol. The first conditional probability is a conditional probability of the symbol given a sub-sequence of the sequence up to the position. The second conditional probability being a conditional probability of the symbol given the sub-sequence.

公开(公告)号：US20190068970A1

公开(公告)日：2019-02-28

申请号：US15707278

申请日：2017-09-18

Applicant: GOOGLE LLC

Inventor： Dake He

IPC: H04N19/13 ,  H04N19/18 ,  H04N19/196 ,  H04N19/159

Abstract: An apparatus for decoding transform coefficients using an alphabet of transform coefficient tokens includes a memory and a processor. The processor is configured to execute instructions stored in the memory to select a first probability distribution corresponding to a first context, select a second probability distribution corresponding to a second context, and, in response to determining that the second probability distribution includes a probability for a transform coefficient token, mix the first probability distribution and the second probability distribution to generate a mixed probability and entropy decode, from an encoded bitstream, the transform coefficient token using the mixed probability. The first probability distribution is defined for all tokens of the alphabet. The second probability distribution is defined over a non-trivial partition of the tokens.

公开(公告)号：US11843513B2

公开(公告)日：2023-12-12

申请号：US17799428

申请日：2020-02-24

Applicant: Google LLC

Inventor： Oleg Golubitsky ,  Pushkarini Hemchandra Agharkar ,  Dake He

IPC: H04L41/142 ,  H04L41/12

CPC classification number: H04L41/142 ,  H04L41/12

Abstract: Systems and methods of enforcing policies in a computer environment for content distribution using pointwise mutual information (PMI) based clustering are provided. The system can maintain a network of nodes representing a plurality of assets. Upon detecting that an asset is associated with a policy label, the system can identify attributes of the asset and compute a PMI score indicating whether nodes of the network sharing the attributes belong to a single content source. Upon determining that the PMI score exceeds a predefined threshold value, the system can identify a cluster of nodes including the nodes sharing the attributes. The system can tag the cluster, for example, as being associated with a content source that is associated with the policy label.

公开(公告)号：US20230325961A1

公开(公告)日：2023-10-12

申请号：US18008544

申请日：2021-06-21

Applicant: Google LLC

Inventor： Dake He ,  Tianhao Zhang ,  Elnaz Barshan Tashnizi ,  Xiyang Luo ,  Huiwen Chang ,  Feng Yang ,  Ryan Matthew Haggarty

IPC: G06T1/00 ,  G06T7/11 ,  G06T3/40

CPC classification number: G06T1/005 ,  G06T7/11 ,  G06T3/40 ,  G06T2201/0083 ,  G06T2207/20081 ,  G06T2201/0065

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining a visually imperceptible or a visually perceptible watermark and outputting a result based on the determination. A watermark decoder receives an input image. The watermark decoder applies a decoder machine learning model to decode a watermarks at different levels of zoom. The water mark decoder determines whether a watermark was decoded to obtain a decoded watermark. The watermark decoder outputs a result based on the determination whether the watermark was decoded through application of the decoder machine learning model to the input image that includes outputting a zoomed output decoded through application of the decoder machine learning model to the input image.

公开(公告)号：US11405618B2

公开(公告)日：2022-08-02

申请号：US17136200

申请日：2020-12-29

Applicant: GOOGLE LLC

Inventor： Dake He

IPC: H04N19/13 ,  H04N19/91 ,  H04N19/18 ,  H04N19/159 ,  H04N19/196 ,  H04N19/124 ,  H04N19/176 ,  H04N19/96 ,  H04N19/129

Abstract: An apparatus for entropy coding a sequence of bits obtains, using a first probability distribution, a first conditional probability for coding a bit at a position within the sequence of bits, the first conditional probability being a conditional probability of the bit having a certain value given that a sub-sequence of the sequence of bits has first respective values; obtains, using a second probability distribution that is different from the first probability distribution, a second conditional probability for coding the bit, the second conditional probability being a conditional probability of the bit having the certain value given that the sub-sequence has second respective values; obtains, using the first conditional probability and the second conditional probability, a mixed probability for coding the bit; and codes the bit using the mixed probability.

公开(公告)号：US11310501B2

公开(公告)日：2022-04-19

申请号：US16868729

申请日：2020-05-07

Applicant: GOOGLE LLC

Inventor： Claudionor Coelho ,  Dake He ,  Aki Kuusela ,  Shan Li

IPC: H04N19/124 ,  H04N19/164 ,  H04N19/176 ,  H04N19/96

Abstract: Encoding an image block using a quantization parameter includes presenting, to an encoder that includes a machine-learning model, the image block and a value derived from the quantization parameter, where the value is a result of a non-linear function using the quantization parameter as input, where the non-linear function relates to a second function used to calculate, using the quantization parameter, a Lagrange multiplier that is used in a rate-distortion calculation, and where the machine-learning model is trained to output mode decision parameters for encoding the image block; obtaining the mode decision parameters from the encoder; and encoding, in a compressed bitstream, the image block using the mode decision parameters.

公开(公告)号：US11025907B2

公开(公告)日：2021-06-01

申请号：US16289149

申请日：2019-02-28

Applicant: GOOGLE LLC

Inventor： Shan Li ,  Claudionor Coelho ,  Aki Kuusela ,  Dake He

IPC: H04N19/107 ,  H04N19/119 ,  H04N19/176 ,  H04N19/96 ,  G06N3/04 ,  G06N3/08

Abstract: Convolutional neural networks (CNN) that determine a mode decision (e.g., block partitioning) for encoding a block include feature extraction layers and multiple classifiers. A non-overlapping convolution operation is performed at a feature extraction layer by setting a stride value equal to a kernel size. The block has a N×N size, and a smallest partition output for the block has a S×S size. Classification layers of each classifier receive feature maps having a feature dimension. An initial classification layer receives the feature maps as an output of a final feature extraction layer. Each classifier infers partition decisions for sub-blocks of size (αS)×(αS) of the block, wherein α is a power of 2 and α=2, . . . , N/S, by applying, at some successive classification layers, a 1×1 kernel to reduce respective feature dimensions; and outputting by a last layer of the classification layers an output corresponding to a N/(αS)×N/(αS)×1 output map.

公开(公告)号：US10798402B2

公开(公告)日：2020-10-06

申请号：US15845161

申请日：2017-12-18

Applicant: GOOGLE LLC

Inventor： Aki Kuusela ,  Dake He

IPC: H04N19/513 ,  H04N19/70 ,  H04N19/14 ,  H04N19/423 ,  H04N19/433 ,  H04N19/593 ,  H04N19/176 ,  H04N19/107

Abstract: Motion estimation or compensation functionality of a hardware component is used to encode or decode key frames and other video frames. The hardware component includes a memory, which may, for example, be a local static random access memory or an external dynamic random access memory. Upon a block of a frame being encoded or decoded, data associated with that block is stored in the memory. That data can then be processed by motion estimation or motion compensation for use in encoding or decoding one or more later blocks within the same frame. The data may, for example, be stored in the memory after operations for reconstruction and loop filtering have been performed. The data stored in the memory may effectively be processed using traditional inter-prediction operations, such as to identify similar video objects within blocks of the same frame.

公开(公告)号：US10609421B2

公开(公告)日：2020-03-31

申请号：US16033582

申请日：2018-07-12

Applicant: Google LLC

Inventor： Aki Kuusela ,  Dake He

IPC: H04N19/91 ,  H04N19/124 ,  H04N19/176 ,  H04N19/625

Abstract: Coding a transform block having transform coefficients is described. A plurality of register arrays is defined to each hold one or more stored values regarding the coding context based on at least one spatial template for a coding context. The register arrays are initialized by setting the stored values to default values, and values for the transform coefficients from the transform block are coded in a reverse scan order. The values for the transform coefficients are indicative of magnitudes of the transform coefficients. For each of one or more transform coefficients, the coding includes determining the coding context using at least some of the stored values from the register arrays, entropy coding a value for the transform coefficient using the coding context, and updating the register arrays subsequent to entropy coding the value for the transform coefficient.

公开(公告)号：US10523968B2

公开(公告)日：2019-12-31

申请号：US15707300

申请日：2017-09-18

Applicant: GOOGLE LLC

Inventor： Dake He

IPC: H04N19/61 ,  H04N19/174 ,  H04N19/64

Abstract: A method for encoding a location of a last non-zero coefficient of a transform block of coefficients includes determining the location of the last non-zero coefficient, the location including a column value and a row value, encoding a non-zero coefficient that is at a coefficient column and a coefficient row of the transform block. The method also includes, if the non-zero coefficient is a first non-zero coefficient to be encoded in the coefficient column, encoding whether the coefficient column is equal to the column value of the last non-zero coefficient, otherwise skipping the encoding. The method also includes, if the non-zero coefficient is a first non-zero coefficient to be encoded in the coefficient row, encoding whether the coefficient row is equal to the row value of the last non-zero coefficient, otherwise skipping the encoding.