Patents by Inventor Johannes Balle
Johannes Balle 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: 20240104786Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for reliably performing data compression and data decompression across a wide variety of hardware and software platforms by using integer neural networks. In one aspect, there is provided a method for entropy encoding data which defines a sequence comprising a plurality of components, the method comprising: for each component of the plurality of components: processing an input comprising: (i) a respective integer representation of each of one or more components of the data which precede the component in the sequence, (ii) an integer representation of one or more respective latent variables characterizing the data, or (iii) both, using an integer neural network to generate data defining a probability distribution over the predetermined set of possible code symbols for the component of the data.Type: ApplicationFiled: November 28, 2023Publication date: March 28, 2024Inventors: Nicholas Johnston, Johannes Balle
-
Publication number: 20240078712Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for compressing and decompressing data. In one aspect, a method comprises: processing data using an encoder neural network to generate a latent representation of the data; processing the latent representation of the data using a hyper-encoder neural network to generate a latent representation of an entropy model; generating an entropy encoded representation of the latent representation of the entropy model; generating an entropy encoded representation of the latent representation of the data using the latent representation of the entropy model; and determining a compressed representation of the data from the entropy encoded representations of: (i) the latent representation of the data and (ii) the latent representation of the entropy model used to entropy encode the latent representation of the data.Type: ApplicationFiled: April 25, 2023Publication date: March 7, 2024Inventors: David Charles Minnen, Saurabh Singh, Johannes Balle, Troy Chinen, Sung Jin Hwang, Nicholas Johnston, George Dan Toderici
-
Patent number: 11907818Abstract: Example aspects of the present disclosure are directed to systems and methods that learn a compressed representation of a machine-learned model (e.g., neural network) via representation of the model parameters within a reparameterization space during training of the model. In particular, the present disclosure describes an end-to-end model weight compression approach that employs a latent-variable data compression method. The model parameters (e.g., weights and biases) are represented in a “latent” or “reparameterization” space, amounting to a reparameterization. In some implementations, this space can be equipped with a learned probability model, which is used first to impose an entropy penalty on the parameter representation during training, and second to compress the representation using arithmetic coding after training. The proposed approach can thus maximize accuracy and model compressibility jointly, in an end-to-end fashion, with the rate-error trade-off specified by a hyperparameter.Type: GrantFiled: February 6, 2023Date of Patent: February 20, 2024Assignee: GOOGLE LLCInventors: Deniz Oktay, Saurabh Singh, Johannes Balle, Abhinav Shrivistava
-
Patent number: 11869221Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for reliably performing data compression and data decompression across a wide variety of hardware and software platforms by using integer neural networks. In one aspect, there is provided a method for entropy encoding data which defines a sequence comprising a plurality of components, the method comprising: for each component of the plurality of components: processing an input comprising: (i) a respective integer representation of each of one or more components of the data which precede the component in the sequence, (ii) an integer representation of one or more respective latent variables characterizing the data, or (iii) both, using an integer neural network to generate data defining a probability distribution over the predetermined set of possible code symbols for the component of the data.Type: GrantFiled: September 18, 2019Date of Patent: January 9, 2024Assignee: Google LLCInventors: Nicholas Johnston, Johannes Balle
-
Publication number: 20230237332Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving, by a neural network (NN), a dataset for generating features from the dataset. A first set of features is computed from the dataset using at least a feature layer of the NN. The first set of features i) is characterized by a measure of informativeness; and ii) is computed such that a size of the first set of features is compressible into a second set of features that is smaller in size than the first set of features and that has a same measure of informativeness as the measure of informativeness of the first set of features. The second set of features if generated from the first set of features using a compression method that compresses the first set of features to generate the second set of features.Type: ApplicationFiled: February 27, 2023Publication date: July 27, 2023Inventors: Abhinav Shrivastava, Saurabh Singh, Johannes Ballé, Sami Ahmad Abu-El-Haija, Nicholas Milo Johnston, George Dan Toderici
-
Publication number: 20230186166Abstract: Example aspects of the present disclosure are directed to systems and methods that learn a compressed representation of a machine-learned model (e.g., neural network) via representation of the model parameters within a reparameterization space during training of the model. In particular, the present disclosure describes an end-to-end model weight compression approach that employs a latent-variable data compression method. The model parameters (e.g., weights and biases) are represented in a “latent” or “reparameterization” space, amounting to a reparameterization. In some implementations, this space can be equipped with a learned probability model, which is used first to impose an entropy penalty on the parameter representation during training, and second to compress the representation using arithmetic coding after training. The proposed approach can thus maximize accuracy and model compressibility jointly, in an end-to-end fashion, with the rate-error trade-off specified by a hyperparameter.Type: ApplicationFiled: February 6, 2023Publication date: June 15, 2023Inventors: Deniz Oktay, Saurabh Singh, Johannes Balle, Abhinav Shrivistava
-
Patent number: 11670010Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for compressing and decompressing data. In one aspect, a method comprises: processing data using an encoder neural network to generate a latent representation of the data; processing the latent representation of the data using a hyper-encoder neural network to generate a latent representation of an entropy model; generating an entropy encoded representation of the latent representation of the entropy model; generating an entropy encoded representation of the latent representation of the data using the latent representation of the entropy model; and determining a compressed representation of the data from the entropy encoded representations of: (i) the latent representation of the data and (ii) the latent representation of the entropy model used to entropy encode the latent representation of the data.Type: GrantFiled: January 19, 2022Date of Patent: June 6, 2023Assignee: Google LLCInventors: David Charles Minnen, Saurabh Singh, Johannes Balle, Troy Chinen, Sung Jin Hwang, Nicholas Johnston, George Dan Toderici
-
Patent number: 11610124Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving, by a neural network (NN), a dataset for generating features from the dataset. A first set of features is computed from the dataset using at least a feature layer of the NN. The first set of features i) is characterized by a measure of informativeness; and ii) is computed such that a size of the first set of features is compressible into a second set of features that is smaller in size than the first set of features and that has a same measure of informativeness as the measure of informativeness of the first set of features. The second set of features if generated from the first set of features using a compression method that compresses the first set of features to generate the second set of features.Type: GrantFiled: October 29, 2019Date of Patent: March 21, 2023Assignee: Google LLCInventors: Abhinav Shrivastava, Saurabh Singh, Johannes Balle, Sami Ahmad Abu-El-Haija, Nicholas Johnston, George Dan Toderici
-
Patent number: 11574232Abstract: Example aspects of the present disclosure are directed to systems and methods that learn a compressed representation of a machine-learned model (e.g., neural network) via representation of the model parameters within a reparameterization space during training of the model. In particular, the present disclosure describes an end-to-end model weight compression approach that employs a latent-variable data compression method. The model parameters (e.g., weights and biases) are represented in a “latent” or “reparameterization” space, amounting to a reparameterization. In some implementations, this space can be equipped with a learned probability model, which is used first to impose an entropy penalty on the parameter representation during training, and second to compress the representation using arithmetic coding after training. The proposed approach can thus maximize accuracy and model compressibility jointly, in an end-to-end fashion, with the rate-error trade-off specified by a hyperparameter.Type: GrantFiled: May 13, 2020Date of Patent: February 7, 2023Assignee: GOOGLE LLCInventors: Deniz Oktay, Saurabh Singh, Johannes Balle, Abhinav Shrivastava
-
Publication number: 20220138991Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for compressing and decompressing data. In one aspect, a method comprises: processing data using an encoder neural network to generate a latent representation of the data; processing the latent representation of the data using a hyper-encoder neural network to generate a latent representation of an entropy model; generating an entropy encoded representation of the latent representation of the entropy model; generating an entropy encoded representation of the latent representation of the data using the latent representation of the entropy model; and determining a compressed representation of the data from the entropy encoded representations of: (i) the latent representation of the data and (ii) the latent representation of the entropy model used to entropy encode the latent representation of the data.Type: ApplicationFiled: January 19, 2022Publication date: May 5, 2022Inventors: David Charles Minnen, Saurabh Singh, Johannes Balle, Troy Chinen, Sung Jin Hwang, Nicholas Johnston, George Dan Toderici
-
Patent number: 11257254Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for compressing and decompressing data. In one aspect, a method comprises: processing data using an encoder neural network to generate a latent representation of the data; processing the latent representation of the data using a hyper-encoder neural network to generate a latent representation of an entropy model; generating an entropy encoded representation of the latent representation of the entropy model; generating an entropy encoded representation of the latent representation of the data using the latent representation of the entropy model; and determining a compressed representation of the data from the entropy encoded representations of: (i) the latent representation of the data and (ii) the latent representation of the entropy model used to entropy encode the latent representation of the data.Type: GrantFiled: July 18, 2019Date of Patent: February 22, 2022Assignee: Google LLCInventors: David Charles Minnen, Saurabh Singh, Johannes Balle, Troy Chinen, Sung Jin Hwang, Nicholas Johnston, George Dan Toderici
-
Publication number: 20210358180Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for reliably performing data compression and data decompression across a wide variety of hardware and software platforms by using integer neural networks. In one aspect, there is provided a method for entropy encoding data which defines a sequence comprising a plurality of components, the method comprising: for each component of the plurality of components: processing an input comprising: (i) a respective integer representation of each of one or more components of the data which precede the component in the sequence, (ii) an integer representation of one or more respective latent variables characterizing the data, or (iii) both, using an integer neural network to generate data defining a probability distribution over the predetermined set of possible code symbols for the component of the data.Type: ApplicationFiled: September 18, 2019Publication date: November 18, 2021Inventors: Nicholas Johnston, Johannes Balle
-
Publication number: 20200364603Abstract: Example aspects of the present disclosure are directed to systems and methods that learn a compressed representation of a machine-learned model (e.g., neural network) via representation of the model parameters within a reparameterization space during training of the model. In particular, the present disclosure describes an end-to-end model weight compression approach that employs a latent-variable data compression method. The model parameters (e.g., weights and biases) are represented in a “latent” or “reparameterization” space, amounting to a reparameterization. In some implementations, this space can be equipped with a learned probability model, which is used first to impose an entropy penalty on the parameter representation during training, and second to compress the representation using arithmetic coding after training. The proposed approach can thus maximize accuracy and model compressibility jointly, in an end-to-end fashion, with the rate-error trade-off specified by a hyperparameter.Type: ApplicationFiled: May 13, 2020Publication date: November 19, 2020Inventors: Deniz Oktay, Saurabh Singh, Johannes Balle, Abhinav Shrivastava
-
Publication number: 20200311548Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving, by a neural network (NN), a dataset for generating features from the dataset. A first set of features is computed from the dataset using at least a feature layer of the NN. The first set of features i) is characterized by a measure of informativeness; and ii) is computed such that a size of the first set of features is compressible into a second set of features that is smaller in size than the first set of features and that has a same measure of informativeness as the measure of informativeness of the first set of features. The second set of features if generated from the first set of features using a compression method that compresses the first set of features to generate the second set of features.Type: ApplicationFiled: October 29, 2019Publication date: October 1, 2020Inventors: Abhinav Shrivastava, Saurabh Singh, Johannes Balle, Sami Ahmad Abu-El-Haija, Nicholas Johnston, George Dan Toderici
-
Publication number: 20200027247Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for compressing and decompressing data. In one aspect, a method comprises: processing data using an encoder neural network to generate a latent representation of the data; processing the latent representation of the data using a hyper-encoder neural network to generate a latent representation of an entropy model; generating an entropy encoded representation of the latent representation of the entropy model; generating an entropy encoded representation of the latent representation of the data using the latent representation of the entropy model; and determining a compressed representation of the data from the entropy encoded representations of: (i) the latent representation of the data and (ii) the latent representation of the entropy model used to entropy encode the latent representation of the data.Type: ApplicationFiled: July 18, 2019Publication date: January 23, 2020Inventors: David Charles Minnen, Saurabh Singh, Johannes Balle, Troy Chinen, Sung Jin Hwang, Nicholas Johnston, George Dan Toderici