Abstract: A method for constructing a free trajectory of a ballistic missile at a specified launch angle includes: setting of an initial state of iterations: based on geodetic coordinates of a launch point and a target point, a launch epoch and the specified launch angle, assuming that the earth does not rotate and a flight time is zero; generating a new flight time in a two-body force model by using known quantities and obtained auxiliary quantities; taking a difference between flight times obtained before and after the iterations as a condition for judging convergence; outputting designed parameters of the ballistic missile after the convergence is reached, or performing a differential correction including the J2 perturbation to improve the precision of the trajectory constructed, and taking a position error of the target point as a convergence condition of the differential correction.

Abstract: Apparatuses, systems, and techniques to facilitate global semi-supervised training of neural networks to perform image segmentation related to diagnosis and management of emerging diseases, such as COVID-19. In at least one embodiment, distributed client training frameworks train one or more client neural networks to perform image segmentation according to a local training data set as well as global neural network data aggregated, by one or more central servers, from each of one or more globally distributed client neural networks.

Abstract: A device (400) for delivering medication to a user includes a circular or elliptical body (410) that includes a mouthpiece (420), a flexible strip (401) of medication, a lever (424), and a mouthpiece cover (491), where the mouthpiece cover is rotatable about the body. An electronics module (120) includes a communication circuit (134), a power supply (126), a sensor system (128), and a switch. The lever is configured to actuate the switch when the lever is moved from a closed position to an open position. The lever is further configured to advance a dose of medication on the flexible strip when moved from the closed position to the open position. The switch switches the electronics module from an off state to an active state when the lever is actuated for a first time by a user. The electronics module is configured to not return to the off state thereafter.

Abstract: Apparatuses, systems, and techniques to generate an optimized neural network architecture. In at least one embodiment, various neural network components are used to generate one or more neural network configurations, and each neural network configuration is trained in order to determine an optimal neural network architecture for a training dataset.

Abstract: Apparatuses, systems, and techniques to train one or more neural networks based, at least in part on, medical imaging data and clinical metadata or inference using one or more neural networks trained as such. In at least one embodiment, one or more circuits to train one or more neural network to predict a treatment for a patient suspected to have or confirmed to have COVID-19 based, at least in part on, medical imaging data and clinical metadata.

Abstract: Comparison logic compares boundaries of features of or more images based, at least in part, on identifying boundaries and indication logic coupled to the comparison logic to indicate whether the boundaries differ by at least a first threshold. The boundaries might comprise a first label mask representing boundaries of objects in an image that are boundaries in a segmentation determined from a segmentation process and a second label mask from a shape evaluation process applied to the first label mask. The indication logic might be configured to compare the first label mask and the second label mask to determine a quality of the segmentation. A neural network might perform the segmentation. Shape evaluation using the first label mask as an input and the second label mask as an output might be performed by a variational autoencoder. A graphical processing unit (GPU) might be used for the segmentation and/or the autoencoder.

Abstract: The present invention provides an imaging probe for imaging mammalian tissues and structures using high resolution imaging, including high frequency ultrasound and optical coherence tomography. The imaging probes structures using high resolution imaging use combined high frequency ultrasound (IVUS) and optical imaging methods such as optical coherence tomography (OCT) and to accurate co-registering of images obtained from ultrasound image signals and optical image signals during scanning a region of interest.

Abstract: Apparatuses, systems, and techniques to train one or more neural networks to generate labels for unsupervised or partially-supervised data. In at least one embodiment, one or more pseudolabels are generated by a training framework based on available weak annotations for an input medical image, and combined with feature information about said input medical image generated by one or more neural networks to generate a label about said input medical image.

Abstract: An electrically conductive adhesive layer includes an adhesive material; a plurality of electrically conductive dendritic first particles dispersed in the adhesive material and having a cumulative 50% particle diameter D50 in a range from about 20 micrometers to about 40 micrometers; and a plurality of electrically conductive substantially planar second particles dispersed in the adhesive material and having a cumulative 50% particle diameter D50 in a range from about 40 micrometers to about 70 micrometers. The adhesive layer has an average thickness in a range from about 15 micrometers to about 35 micrometers, an electrical resistance in a thickness direction of less than about 30 milliohms, and a peel strength of at least 0.1 N/mm from a stainless steel surface after a dwell time of about 20 minutes at 22° C.

Abstract: Apparatuses, systems, and techniques to train neural networks to perform classification. In at least one embodiment, one or more neural networks are trained to perform classification based on, for example, using one or more compressed representations of one or more class labels, where the one or more compressed representations have fewer bits than a representation of the one or more class labels.

Abstract: A water spraying device comprises a housing comprising a mechanism mounting cavity, an effuser mounting cavity located at a front portion of the mechanism mounting cavity, and a water pipe accommodating cavity located at an upper portion of the mechanism mounting cavity and communicated with the effuser mounting cavity. The device also comprises an effuser comprising a nozzle and slidably arranged in the effuser mounting cavity; a water supply hose accommodated in the water pipe accommodating cavity and configured to supply water and move in the water pipe accommodating cavity; a driving mechanism mounted in the mechanism mounting cavity and configured to drive the effuser to move; and a transmission mechanism connecting the driving mechanism to the effuser. The housing comprises a housing water inlet and the effuser comprises an effuser water inlet. The water supply hose is connected between the housing water inlet and the effuser water inlet.

Abstract: A method for customizing facial expressions of a user includes: obtaining a RGB-D image sequence, performing non-rigid registration on a three-dimensional (3D) face template model with a depth map and face feature points corresponding to the RGB-D image sequence, inputting each vertex in the non-rigid registration result into the depth map to generate a set of deformation data, and deforming the 3D face template model based on the set of deformation data; reconstructing face details in the RGB-D image sequence, and generating a 3D neutral face model based on the deformed 3D face template model and the reconstructed 3D face template model; processing the 3D neutral face model and a face hybrid template to generate a face hybrid model; tracking the face in the RGB-D image sequence to generate a face tracking result; and updating the face hybrid model based on the face tracking result and customizing the facial expressions.

Abstract: In at least one embodiment, an object detection system uses a neural network to identify and/or locate a set of organs in a medical image. In at least one embodiment, when training to identify and/or locate a particular organ, a subset of incompletely-labeled training images is used that excludes training images for which labels associated with particular organ are unavailable.

Abstract: Apparatuses, systems, and techniques to select labels of training images to train a network. In at least one embodiment, one or more labels of training images are selected to train a network.

Abstract: Apparatuses, systems, and techniques to select a nueral network architecture from a plurality of neural networs in a federated learning (FL) settng. In at least one embodiment, a neural network is trained by combining training resutls from different FL computing systesms, where each of the different FL computing systems, for example, trains different portions of the nerual network.

Abstract: An electrically conductive adhesive layer is described. The adhesive layer includes an adhesive material and pluralities of electrically conductive at least first and second particles. The adhesive layer may have a thickness less than about 35 micrometers and an electrical resistance in the thickness direction of less than about 30 milliohms. A total volume of the pluralities of particles may be greater than 40% of a total volume of the adhesive layer. The first and second particles may have different shapes.

Abstract: A valve actuation system comprising a valve actuation motion source configured to provide a main event valve actuation motion to at least one engine valve via a main motion load path that comprises at least one valve train component. The valve actuation system further includes a lost motion component arranged within a first valve train component in the main motion load path, the lost motion component being controllable to operate in a motion conveying state or a motion absorbing state. The valve actuation system also comprises a high lift transfer component arranged in the main motion load path, with the high lift transfer component being configured to permit the main motion load path to convey at least a high lift portion of the main event valve actuation motion when the lost motion component is in the motion absorbing state.

Abstract: Apparatuses, systems, and techniques are presented to predict segmentations for objects in images. In at least one embodiment, a neural network is trained to determine one or more segmentation masks corresponding to one or more objects of one or more digital images based, at least in part, on one or more boundary regions of the one or more objects.

Abstract: Apparatuses, systems, and techniques are presented to predict annotations for objects in images. In at least one embodiment, one or more annotations corresponding to one or more objects within one or more images are generated based, at least in part, on one or more neural networks iteratively trained using the one or more images.

Abstract: A method of manufacturing a jacketed metal line is detailed herein. The method of manufacturing a jacketed metal line can include roughening an outer surface of a metal core of the line. An insulating polymer layer can be applied to the metal core, and the insulating polymer layer can include a reinforcing additive comprising: graphite, carbon, glass, aramid, short-fiber filled PolyEtherEtherKetone, mircron-sized polytetrafluoroethylene, or combinations thereof. The roughened metal core can then be exposed a heat source for at least partially melting the polymer layer; and the partially melted polymer layer and insulated roughened metal core can be ran through a set of shaping rollers.