Patents by Inventor Gavriel State
Gavriel State 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: 11995883Abstract: Approaches are presented for training and using scene graph generators for transfer learning. A scene graph generation technique can decompose a domain gap into individual types of discrepancies, such as may relate to appearance, label, and prediction discrepancies. These discrepancies can be reduced, at least in part, by aligning the corresponding latent and output distributions using one or more gradient reversal layers (GRLs). Label discrepancies can be addressed using self-pseudo-statistics collected from target data. Pseudo statistic-based self-learning and adversarial techniques can be used to manage these discrepancies without the need for costly supervision from a real-world dataset.Type: GrantFiled: February 2, 2023Date of Patent: May 28, 2024Assignee: Nvidia CorporationInventors: Aayush Prakash, Shoubhik Debnath, Jean-Francois Lafleche, Eric Cameracci, Gavriel State, Marc Teva Law
-
Publication number: 20240100694Abstract: Systems techniques to control a robot are described herein. In at least one embodiment, a machine learning model for controlling a robot is trained based at least on one or more population-based training operations or one or more reinforcement learning operations. Once trained, the machine learning model can be deployed and used to control a robot to perform a task.Type: ApplicationFiled: June 7, 2023Publication date: March 28, 2024Inventors: Ankur HANDA, Gavriel STATE, Arthur David ALLSHIRE, Victor MAKOVIICHUK, Aleksei Vladimirovich PETRENKO
-
Publication number: 20240095527Abstract: Systems and techniques are described related to training one or more machine learning models for use in control of a robot. In at least one embodiment, one or more machine learning models are trained based at least on simulations of the robot and renderings of such simulations—which may be performed using one or more ray tracing algorithms, operations, or techniques.Type: ApplicationFiled: August 10, 2023Publication date: March 21, 2024Inventors: Ankur HANDA, Gavriel STATE, Arthur David ALLSHIRE, Dieter FOX, Jean-Francois Victor LAFLECHE, Jingzhou LIU, Viktor MAKOVIICHUK, Yashraj Shyam NARANG, Aleksei Vladimirovich PETRENKO, Ritvik SINGH, Balakumar SUNDARALINGAM, Karl VAN WYK, Alexander ZHURKEVICH
-
Publication number: 20230321822Abstract: One embodiment of a method for controlling a robot includes performing a plurality of simulations of a robot interacting with one or more objects represented by one or more signed distance functions (SDFs), where performing the plurality of simulations comprises reducing a number of contacts between the one or more objects that are being simulated, and updating one or more parameters of a machine learning model based on the plurality of simulations to generate a trained machine learning model.Type: ApplicationFiled: December 2, 2022Publication date: October 12, 2023Inventors: Yashraj Shyam NARANG, Kier STOREY, Iretiayo AKINOLA, Dieter FOX, Kelly GUO, Ankur HANDA, Fengyun LU, Miles MACKLIN, Adam MORAVANSZKY, Philipp REIST, Gavriel STATE, Lukasz WAWRZYNIAK
-
Publication number: 20230177826Abstract: Approaches are presented for training and using scene graph generators for transfer learning. A scene graph generation technique can decompose a domain gap into individual types of discrepancies, such as may relate to appearance, label, and prediction discrepancies. These discrepancies can be reduced, at least in part, by aligning the corresponding latent and output distributions using one or more gradient reversal layers (GRLs). Label discrepancies can be addressed using self-pseudo-statistics collected from target data. Pseudo statistic-based self-learning and adversarial techniques can be used to manage these discrepancies without the need for costly supervision from a real-world dataset.Type: ApplicationFiled: February 2, 2023Publication date: June 8, 2023Inventors: Aayush Prakash, Shoubhik Debnath, Jean-Francois Lafleche, Eric Cameracci, Gavriel State, Marc Teva Law
-
Patent number: 11574155Abstract: Approaches are presented for training and using scene graph generators for transfer learning. A scene graph generation technique can decompose a domain gap into individual types of discrepancies, such as may relate to appearance, label, and prediction discrepancies. These discrepancies can be reduced, at least in part, by aligning the corresponding latent and output distributions using one or more gradient reversal layers (GRLs). Label discrepancies can be addressed using self-pseudo-statistics collected from target data. Pseudo statistic-based self-learning and adversarial techniques can be used to manage these discrepancies without the need for costly supervision from a real-world dataset.Type: GrantFiled: April 9, 2021Date of Patent: February 7, 2023Assignee: Nvidia CorporationInventors: Aayush Prakash, Shoubhik Debnath, Jean-Francois Lafleche, Eric Cameracci, Gavriel State, Marc Teva Law
-
Publication number: 20220230376Abstract: Animation can be generated with a high perceptive quality by utilizing a trained neural network that takes as input a current state of a virtual character to be animated and predict how this character would appear in one or more subsequent frames. Such a process can be performed recursively to generate the data for these frames. During training, each frame of a generated sequence can be predicted from a result for a previous frame, and this generated sequence can be compared with a ground truth sequence using a generative network. Differences between the ground truth and generated animation sequences can be minimized, whereby a specific objective function does not need to be manually defined. Minimizing differences between the generated animation sequences and ground truth sequences during training improves the quality of network predictions for single frames at inference time.Type: ApplicationFiled: May 15, 2020Publication date: July 21, 2022Inventors: Artem Rozantsev, Marco Foco, Gavriel State
-
Publication number: 20210374489Abstract: Approaches are presented for training and using scene graph generators for transfer learning. A scene graph generation technique can decompose a domain gap into individual types of discrepancies, such as may relate to appearance, label, and prediction discrepancies. These discrepancies can be reduced, at least in part, by aligning the corresponding latent and output distributions using one or more gradient reversal layers (GRLs). Label discrepancies can be addressed using self-pseudo-statistics collected from target data. Pseudo statistic-based self-learning and adversarial techniques can be used to manage these discrepancies without the need for costly supervision from a real-world dataset.Type: ApplicationFiled: April 9, 2021Publication date: December 2, 2021Inventors: Aayush Prakash, Shoubhik Debnath, Jean-Francois Lafleche, Eric Cameracci, Gavriel State, Marc Teva Law
-
Patent number: 9875138Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: GrantFiled: August 1, 2016Date of Patent: January 23, 2018Assignee: Google LLCInventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Publication number: 20170132038Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: ApplicationFiled: August 1, 2016Publication date: May 11, 2017Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Patent number: 9477452Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: GrantFiled: February 25, 2015Date of Patent: October 25, 2016Assignee: Google Inc.Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Patent number: 9436451Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: GrantFiled: November 13, 2015Date of Patent: September 6, 2016Assignee: GOOGLE INC.Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Patent number: 9430202Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: GrantFiled: November 13, 2015Date of Patent: August 30, 2016Assignee: GOOGLE INC.Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Publication number: 20160070552Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: ApplicationFiled: November 13, 2015Publication date: March 10, 2016Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Publication number: 20160071305Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: ApplicationFiled: November 13, 2015Publication date: March 10, 2016Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Publication number: 20150169305Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: ApplicationFiled: February 25, 2015Publication date: June 18, 2015Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Patent number: 9019283Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: GrantFiled: August 29, 2012Date of Patent: April 28, 2015Assignee: Transgaming Inc.Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Publication number: 20120320051Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: ApplicationFiled: August 29, 2012Publication date: December 20, 2012Applicant: TRANSGAMING TECHNOLOGIES INC.Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Patent number: 8284206Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: GrantFiled: March 14, 2007Date of Patent: October 9, 2012Assignee: Transgaming, Inc.Inventors: Gavriel State, Nicolas Capens, Luther Johnson
-
Publication number: 20070220525Abstract: A software engine for decomposing work to be done into tasks, and distributing the tasks to multiple, independent CPUs for execution is described. The engine utilizes dynamic code generation, with run-time specialization of variables, to achieve high performance. Problems are decomposed according to methods that enhance parallel CPU operation, and provide better opportunities for specialization and optimization of dynamically generated code. A specific application of this engine, a software three dimensional (3D) graphical image renderer, is described.Type: ApplicationFiled: March 14, 2007Publication date: September 20, 2007Inventors: Gavriel STATE, Nicolas Capens, Luther Johnson