Abstract: There is described a system for providing an interface for product personalization by generating or modifying garment patterns and associated code files. The system processes the extracted measurement attributes and user preferences using generative design models and customization data from a control panel of tools at the interface. The system involves an interface to display a visualization of a garment and uses output of a modeling system to support the visualization. The interface has the control panel of tools to customize, modify or generate garment patterns and associated code files.

Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.

Abstract: A system and method for carbon dioxide capture/storage from exhaust utilizing a carbon dioxide separation membrane which may be prepared in the form of a monolithic structure. The captured carbon dioxide may also be stored in a fluid state as supercritical CO2. An integrated fuel delivery and carbon dioxide unloading system is also disclosed, to remove carbon dioxide from a vehicle for sequestration or other industrial purposes.

Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.

Abstract: Method for obtaining a directional survey in a rotating downhole component include acquiring and generating, while rotating, sets of raw data having 3-axis magnetic field data and 3-axis gravity field data. A set of rotationally-invariant data is obtained for each of the sets of raw data to generate a first number of sets of rotationally-invariant data. An earth property value is calculated from each of the sets. An accuracy indicator is estimated for each of the sets using a respective earth property value, an earth property reference value, and an error model, to generate a plurality of accuracy indicators. A set of mean values is determined using the plurality of sets of rotationally-invariant data using a second number of sets of rotationally-invariant data of the plurality of sets of rotationally-invariant data. The directional survey is estimated and used to control the downhole component.

Abstract: In various examples, a deep neural network (DNN) is trained—using image data alone—to accurately predict distances to objects, obstacles, and/or a detected free-space boundary. The DNN may be trained with ground truth data that is generated using sensor data representative of motion of an ego-vehicle and/or sensor data from any number of depth predicting sensors—such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. The DNN may be trained using two or more loss functions each corresponding to a particular portion of the environment that depth is predicted for, such that—in deployment—more accurate depth estimates for objects, obstacles, and/or the detected free-space boundary are computed by the DNN. In some embodiments, a sampling algorithm may be used to sample depth values corresponding to an input resolution of the DNN from a predicted depth map of the DNN at an output resolution of the DNN.

Abstract: Method and apparatus for predicting an action to be taken by a user of a mobile device, comprising: generating a machine learning (ML) model using historical training data comprising (a) output training data that includes a scheduled action on a calendar and (b) input training data comprising historical usage data of the mobile device by the user; receiving real-time usage data of the mobile device, the real-time usage data indicating at least one of (a) an emotional state of the user, wherein the emotional state is indicative of autism spectrum disorder, or (b) indicators of the environment of the user; and evaluating data comprising the real-time usage data of the mobile device using the ML model to output a suggested action to be performed by the user, wherein the suggested action mitigates a negative impact on the user of a symptom of the autism spectrum disorder.

Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.

Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.

Abstract: In various examples, a deep neural network (DNN) is trained—using image data alone—to accurately predict distances to objects, obstacles, and/or a detected free-space boundary. The DNN may be trained with ground truth data that is generated using sensor data representative of motion of an ego-vehicle and/or sensor data from any number of depth predicting sensors—such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. The DNN may be trained using two or more loss functions each corresponding to a particular portion of the environment that depth is predicted for, such that—in deployment—more accurate depth estimates for objects, obstacles, and/or the detected free-space boundary are computed by the DNN. In some embodiments, a sampling algorithm may be used to sample depth values corresponding to an input resolution of the DNN from a predicted depth map of the DNN at an output resolution of the DNN.

Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.

Abstract: Embodiments relate generally to computer software and computing devices, and more particularly, to a system, an apparatus and a method configured to segregate data at an interface of a computing device to facilitate on-line electronic payment transactions. In one embodiment, a method includes presenting fields configured to accept a first type of data and to accept a second type of data for an on-line electronic payment transaction. The method includes generating an initialization signal for transmission to an isolated data management system to initialize a portion of a memory associated with first type of data, responsive to an interaction with a field, receiving data from the field, and generating a save signal to save the data from the field in a portion of the memory. This can be responsive to a second interaction with the field.

Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.

Abstract: Embodiments relate generally to computer software and computing devices, and more particularly, to a system, an apparatus and a method configured to segregate data at an interface of a computing device to facilitate on-line electronic payment transactions. In one embodiment, a method includes presenting fields configured to accept a first type of data and to accept a second type of data for an on-line electronic payment transaction. The method includes generating an initialization signal for transmission to an isolated data management system to initialize a portion of a memory associated with first type of data, responsive to an interaction with a field, receiving data from the field, and generating a save signal to save the data from the field in a portion of the memory. This can be responsive to a second interaction with the field.

Abstract: In various examples, a deep neural network (DNN) is trained to accurately predict, in deployment, distances to objects and obstacles using image data alone. The DNN may be trained with ground truth data that is generated and encoded using sensor data from any number of depth predicting sensors, such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. Camera adaptation algorithms may be used in various embodiments to adapt the DNN for use with image data generated by cameras with varying parameters—such as varying fields of view. In some examples, a post-processing safety bounds operation may be executed on the predictions of the DNN to ensure that the predictions fall within a safety-permissible range.