Patents by Inventor Harsimran Singh
Harsimran Singh 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: 20250005593Abstract: A method for bulk data validation for draft-based SAP Fiori applications includes determining a uniform resource location (URL) of each of at least a draft business entity for a draft-based SAP Fiori application by identifying at least a location call for each of the at least a draft business entity in at least an OData batch call for the draft-based SAP Fiori application. Service metadata information is obtained from an SAP server and input parameters for each draft business entity are identified. Data type for each input parameter is determined using the service metadata information. The input parameters are sent to a user interface. The input parameters are updated with user input. The method further includes validating the updated input parameters for each of the at least a draft business entity by replaying the at least an OData batch call with the updated at least a location call.Type: ApplicationFiled: June 30, 2023Publication date: January 2, 2025Inventors: Harsimran Singh Dhami, Harminder Singh
-
Publication number: 20240419968Abstract: A model training and implementation pipeline trains models for individual embedded systems. The pipeline iterates through multiple models and estimates the performance of the models. During a model generation stage, the pipeline translates the description of the model together with the model parameters into an intermediate representation in a language that is compatible with a virtual machine. The intermediate representation is agnostic or independent to the configuration of the target platform. During a model performance estimation stage, the pipeline evaluates the performance of the models without training the models. Based on the analysis of the performance of the untrained models, a subset of models is selected. The selected models are then trained and the performance of the trained models are analyzed. Based on the analysis of the performance of the trained models, a single model is selected for deployment to the target platform.Type: ApplicationFiled: August 30, 2024Publication date: December 19, 2024Applicant: Tesla, Inc.Inventors: Harsimran Singh Sidhu, Paras Jagdish Jain, Daniel Paden Tomasello, Forrest Nelson Iandola
-
Publication number: 20240346816Abstract: An autonomous control system generates synthetic data that reflect simulated environments. Specifically, the synthetic data is a representation of sensor data of the simulated environment from the perspective of one or more sensors. The system generates synthetic data by introducing one or more simulated modifications to sensor data captured by the sensors or by simulating the sensor data for a virtual environment. The autonomous control system uses the synthetic data to train computer models for various detection and control algorithms. In general, this allows autonomous control systems to augment training data to improve performance of computer models, simulate scenarios that are not included in existing training data, and/or train computer models that remove unwanted effects or occlusions from sensor data of the environment.Type: ApplicationFiled: June 24, 2024Publication date: October 17, 2024Inventors: Forrest Nelson Iandola, Donald Benton MacMillen, Anting Shen, Harsimran Singh Sidhu, Paras Jagdish Jain
-
Publication number: 20240296330Abstract: A neural network architecture is used that reduces the processing load of implementing the neural network. This network architecture may thus be used for reduced-bit processing devices. The architecture may limit the number of bits used for processing and reduce processing to prevent data overflow at individual calculations of the neural network. To implement this architecture, the number of bits used to represent inputs at levels of the network and the related filter masks may also be modified to ensure the number of bits of the output does not overflow the resulting capacity of the reduced-bit processor. To additionally reduce the load for such a network, the network may implement a “starconv” structure that permits the incorporation of nearby nodes in a layer to balance processing requirements and permit the network to learn from context of other nodes.Type: ApplicationFiled: May 14, 2024Publication date: September 5, 2024Inventors: Forrest Nelson Iandola, Harsimran Singh Sidhu, Yiqi Hou
-
Patent number: 12079723Abstract: A model training and implementation pipeline trains models for individual embedded systems. The pipeline iterates through multiple models and estimates the performance of the models. During a model generation stage, the pipeline translates the description of the model together with the model parameters into an intermediate representation in a language that is compatible with a virtual machine. The intermediate representation is agnostic or independent to the configuration of the target platform. During a model performance estimation stage, the pipeline evaluates the performance of the models without training the models. Based on the analysis of the performance of the untrained models, a subset of models is selected. The selected models are then trained and the performance of the trained models are analyzed. Based on the analysis of the performance of the trained models, a single model is selected for deployment to the target platform.Type: GrantFiled: March 14, 2023Date of Patent: September 3, 2024Assignee: Tesla, Inc.Inventors: Harsimran Singh Sidhu, Paras Jagdish Jain, Daniel Paden Tomasello, Forrest Nelson Iandola
-
Patent number: 12020476Abstract: An autonomous control system generates synthetic data that reflect simulated environments. Specifically, the synthetic data is a representation of sensor data of the simulated environment from the perspective of one or more sensors. The system generates synthetic data by introducing one or more simulated modifications to sensor data captured by the sensors or by simulating the sensor data for a virtual environment. The autonomous control system uses the synthetic data to train computer models for various detection and control algorithms. In general, this allows autonomous control systems to augment training data to improve performance of computer models, simulate scenarios that are not included in existing training data, and/or train computer models that remove unwanted effects or occlusions from sensor data of the environment.Type: GrantFiled: October 28, 2022Date of Patent: June 25, 2024Assignee: Tesla, Inc.Inventors: Forrest Nelson Iandola, Donald Benton MacMillen, Anting Shen, Harsimran Singh Sidhu, Paras Jagdish Jain
-
Publication number: 20240177455Abstract: Systems and methods for training machine models with augmented data. An example method includes identifying a set of images captured by a set of cameras while affixed to one or more image collection systems. For each image in the set of images, a training output for the image is identified. For one or more images in the set of images, an augmented image for a set of augmented images is generated. Generating an augmented image includes modifying the image with an image manipulation function that maintains camera properties of the image. The augmented training image is associated with the training output of the image. A set of parameters of the predictive computer model are trained to predict the training output based on an image training set including the images and the set of augmented images.Type: ApplicationFiled: February 5, 2024Publication date: May 30, 2024Applicant: Tesla, Inc.Inventors: Matthew John Cooper, Paras Jagdish Jain, Harsimran Singh Sidhu
-
Patent number: 11983630Abstract: A neural network architecture is used that reduces the processing load of implementing the neural network. This network architecture may thus be used for reduced-bit processing devices. The architecture may limit the number of bits used for processing and reduce processing to prevent data overflow at individual calculations of the neural network. To implement this architecture, the number of bits used to represent inputs at levels of the network and the related filter masks may also be modified to ensure the number of bits of the output does not overflow the resulting capacity of the reduced-bit processor. To additionally reduce the load for such a network, the network may implement a “starconv” structure that permits the incorporation of nearby nodes in a layer to balance processing requirements and permit the network to learn from context of other nodes.Type: GrantFiled: January 19, 2023Date of Patent: May 14, 2024Assignee: Tesla, Inc.Inventors: Forrest Nelson Iandola, Harsimran Singh Sidhu, Yiqi Hou
-
Patent number: 11893774Abstract: Systems and methods for training machine models with augmented data. An example method includes identifying a set of images captured by a set of cameras while affixed to one or more image collection systems. For each image in the set of images, a training output for the image is identified. For one or more images in the set of images, an augmented image for a set of augmented images is generated. Generating an augmented image includes modifying the image with an image manipulation function that maintains camera properties of the image. The augmented training image is associated with the training output of the image. A set of parameters of the predictive computer model are trained to predict the training output based on an image training set including the images and the set of augmented images.Type: GrantFiled: December 14, 2021Date of Patent: February 6, 2024Assignee: Tesla, Inc.Inventors: Matthew John Cooper, Paras Jagdish Jain, Harsimran Singh Sidhu
-
Publication number: 20240000192Abstract: A method for providing a user interface for HTML SAP applications includes identifying, by a user device and among network traffic generated for at least an SAP application, representational state transfer application programming interface (REST API) calls sent to a server. The method further includes identifying, by the user device, user-filled keys in the identified REST API calls. The method further includes displaying, by a user interface on the user device, the user-filled keys. The method further includes receiving user input for the user-filled keys on the user interface. The method further includes updating the REST API calls based on the received user input and executing the updated REST API calls. The method further includes displaying responses to the REST API calls from the server on the user interface.Type: ApplicationFiled: June 30, 2023Publication date: January 4, 2024Inventor: Harsimran Singh Dhami
-
Publication number: 20230289599Abstract: A model training and implementation pipeline trains models for individual embedded systems. The pipeline iterates through multiple models and estimates the performance of the models. During a model generation stage, the pipeline translates the description of the model together with the model parameters into an intermediate representation in a language that is compatible with a virtual machine. The intermediate representation is agnostic or independent to the configuration of the target platform. During a model performance estimation stage, the pipeline evaluates the performance of the models without training the models. Based on the analysis of the performance of the untrained models, a subset of models is selected. The selected models are then trained and the performance of the trained models are analyzed. Based on the analysis of the performance of the trained models, a single model is selected for deployment to the target platform.Type: ApplicationFiled: March 14, 2023Publication date: September 14, 2023Inventors: Harsimran Singh Sidhu, Paras Jagdish Jain, Daniel Paden Tomasello, Forrest Nelson Iandola
-
Publication number: 20230237331Abstract: A neural network architecture is used that reduces the processing load of implementing the neural network. This network architecture may thus be used for reduced-bit processing devices. The architecture may limit the number of bits used for processing and reduce processing to prevent data overflow at individual calculations of the neural network. To implement this architecture, the number of bits used to represent inputs at levels of the network and the related filter masks may also be modified to ensure the number of bits of the output does not overflow the resulting capacity of the reduced-bit processor. To additionally reduce the load for such a network, the network may implement a “starconv” structure that permits the incorporation of nearby nodes in a layer to balance processing requirements and permit the network to learn from context of other nodes.Type: ApplicationFiled: January 19, 2023Publication date: July 27, 2023Inventors: Forrest Nelson Iandola, Harsimran Singh Sidhu, Yiqi Hou
-
Publication number: 20230177819Abstract: An autonomous control system generates synthetic data that reflect simulated environments. Specifically, the synthetic data is a representation of sensor data of the simulated environment from the perspective of one or more sensors. The system generates synthetic data by introducing one or more simulated modifications to sensor data captured by the sensors or by simulating the sensor data for a virtual environment. The autonomous control system uses the synthetic data to train computer models for various detection and control algorithms. In general, this allows autonomous control systems to augment training data to improve performance of computer models, simulate scenarios that are not included in existing training data, and/or train computer models that remove unwanted effects or occlusions from sensor data of the environment.Type: ApplicationFiled: October 28, 2022Publication date: June 8, 2023Inventors: Forrest Nelson Iandola, Donald Benton MacMillen, Anting Shen, Harsimran Singh Sidhu, Paras Jagdish Jain
-
Patent number: 11636333Abstract: A model training and implementation pipeline trains models for individual embedded systems. The pipeline iterates through multiple models and estimates the performance of the models. During a model generation stage, the pipeline translates the description of the model together with the model parameters into an intermediate representation in a language that is compatible with a virtual machine. The intermediate representation is agnostic or independent to the configuration of the target platform. During a model performance estimation stage, the pipeline evaluates the performance of the models without training the models. Based on the analysis of the performance of the untrained models, a subset of models is selected. The selected models are then trained and the performance of the trained models are analyzed. Based on the analysis of the performance of the trained models, a single model is selected for deployment to the target platform.Type: GrantFiled: July 25, 2019Date of Patent: April 25, 2023Assignee: Tesla, Inc.Inventors: Harsimran Singh Sidhu, Paras Jagdish Jain, Daniel Paden Tomasello, Forrest Nelson Iandola
-
Patent number: 11562231Abstract: A neural network architecture is used that reduces the processing load of implementing the neural network. This network architecture may thus be used for reduced-bit processing devices. The architecture may limit the number of bits used for processing and reduce processing to prevent data overflow at individual calculations of the neural network. To implement this architecture, the number of bits used to represent inputs at levels of the network and the related filter masks may also be modified to ensure the number of bits of the output does not overflow the resulting capacity of the reduced-bit processor. To additionally reduce the load for such a network, the network may implement a “starconv” structure that permits the incorporation of nearby nodes in a layer to balance processing requirements and permit the network to learn from context of other nodes.Type: GrantFiled: September 3, 2019Date of Patent: January 24, 2023Assignee: Tesla, Inc.Inventors: Forrest Nelson Iandola, Harsimran Singh Sidhu, Yiqi Hou
-
Patent number: 11487288Abstract: An autonomous control system generates synthetic data that reflect simulated environments. Specifically, the synthetic data is a representation of sensor data of the simulated environment from the perspective of one or more sensors. The system generates synthetic data by introducing one or more simulated modifications to sensor data captured by the sensors or by simulating the sensor data for a virtual environment. The autonomous control system uses the synthetic data to train computer models for various detection and control algorithms. In general, this allows autonomous control systems to augment training data to improve performance of computer models, simulate scenarios that are not included in existing training data, and/or train computer models that remove unwanted effects or occlusions from sensor data of the environment.Type: GrantFiled: June 8, 2020Date of Patent: November 1, 2022Assignee: Tesla, Inc.Inventors: Forrest Nelson Iandola, Donald Benton MacMillen, Anting Shen, Harsimran Singh Sidhu, Paras Jagdish Jain
-
Publication number: 20220108130Abstract: Systems and methods for training machine models with augmented data. An example method includes identifying a set of images captured by a set of cameras while affixed to one or more image collection systems. For each image in the set of images, a training output for the image is identified. For one or more images in the set of images, an augmented image for a set of augmented images is generated. Generating an augmented image includes modifying the image with an image manipulation function that maintains camera properties of the image. The augmented training image is associated with the training output of the image. A set of parameters of the predictive computer model are trained to predict the training output based on an image training set including the images and the set of augmented images.Type: ApplicationFiled: December 14, 2021Publication date: April 7, 2022Inventors: Matthew John Cooper, Paras Jagdish Jain, Harsimran Singh Sidhu
-
Publication number: 20220043449Abstract: An autonomous control system combines sensor data from multiple sensors to simulate sensor data from high-capacity sensors. The sensor data contains information related to physical environments surrounding vehicles for autonomous guidance. For example, the sensor data may be in the form of images that visually capture scenes of the surrounding environment, geo-location of the vehicles, and the like. The autonomous control system simulates high-capacity sensor data of the physical environment from replacement sensors that may each have lower capacity than high-capacity sensors. The high-capacity sensor data may be simulated via one or more neural network models. The autonomous control system performs various detection and control algorithms on the simulated sensor data to guide the vehicle autonomously.Type: ApplicationFiled: October 22, 2021Publication date: February 10, 2022Inventors: Forrest Nelson Iandola, Donald Benton MacMillen, Anting Shen, Harsimran Singh Sidhu, Daniel Paden Tomasello, Rohan Nandkumar Phadte, Paras Jagdish Jian
-
Patent number: 11205093Abstract: Systems and methods for training machine models with augmented data. An example method includes identifying a set of images captured by a set of cameras while affixed to one or more image collection systems. For each image in the set of images, a training output for the image is identified. For one or more images in the set of images, an augmented image for a set of augmented images is generated. Generating an augmented image includes modifying the image with an image manipulation function that maintains camera properties of the image. The augmented training image is associated with the training output of the image. A set of parameters of the predictive computer model are trained to predict the training output based on an image training set including the images and the set of augmented images.Type: GrantFiled: October 10, 2019Date of Patent: December 21, 2021Assignee: Tesla, Inc.Inventors: Matthew John Cooper, Paras Jagdish Jain, Harsimran Singh Sidhu
-
Patent number: 11157014Abstract: An autonomous control system combines sensor data from multiple sensors to simulate sensor data from high-capacity sensors. The sensor data contains information related to physical environments surrounding vehicles for autonomous guidance. For example, the sensor data may be in the form of images that visually capture scenes of the surrounding environment, geo-location of the vehicles, and the like. The autonomous control system simulates high-capacity sensor data of the physical environment from replacement sensors that may each have lower capacity than high-capacity sensors. The high-capacity sensor data may be simulated via one or more neural network models. The autonomous control system performs various detection and control algorithms on the simulated sensor data to guide the vehicle autonomously.Type: GrantFiled: December 27, 2017Date of Patent: October 26, 2021Assignee: Tesla, Inc.Inventors: Forrest Nelson Iandola, Donald Benton MacMillen, Anting Shen, Harsimran Singh Sidhu, Daniel Paden Tomasello, Rohan Nandkumar Phadte, Paras Jagdish Jain