Abstract: Discussed herein are devices, systems, and methods for classification using a clustering autoencoder. A method can include obtaining content to be classified by the DNN classifier, and operating the DNN classifier to determine a classification of the received content, the DNN classifier including a clustering classification layer that clusters based on a latent feature vector representation of the content, the classification corresponding to one or more clusters that are closest to the latent feature vector providing the classification and a corresponding confidence.

Abstract: Techniques for multi-factor pathfinding are disclosed. A system receives multiple data sets that represent multiple physical properties associated with multiple three-dimensional coordinates of a physical environment. The system generates multiple weighted data layers corresponding to the physical properties, at least by applying weights to the data sets. The system generates a multi-factor costmap for the three-dimensional environment, based at least on the weighted data layers. The system determines a path through the three-dimensional environment, based at least on the multi-factor costmap.

Abstract: Subject matter regards improving machine learning techniques using informed pseudolabels. A method can include receiving previously assigned labels indicating an expected classification for data, the labels having a specified uncertainty, generating respective pseudolabels for the data based on the previously assigned labels, the data, a class vector determined by an ML model, and a noise model indicating, based on the specified uncertainty, a likelihood of the previously assigned label given the class, and substituting the pseudolabels for the previously assigned labels in a next epoch of training the ML model.

Abstract: An apparatus for a wireless power transfer (“WPT”) pad heat management system includes a ferrite structure positioned adjacent to a coil configured to wirelessly transfer power. The apparatus includes a plurality of heat spreaders positioned along a length of a component of the ferrite structure. Each of the plurality of heat spreaders is non-metallic. The apparatus includes a trough shaped to surround at least a portion of each of the plurality of heat spreaders, wherein the trough is non-metallic. The apparatus includes a phase change material (“PCM”) in the trough where at least a portion of the heat spreaders extend into the PCM. The ferrite structure, coil, plurality of heat spreaders, trough and PCM are encased in a solid material, and each of the plurality of heat spreaders comprises a material that transfers heat from the component of the ferrite structure to the PCM.

Abstract: Discussed herein are devices, systems, and methods for classification using a clustering autoencoder. A method can include obtaining content to be classified by the DNN classifier, and operating the DNN classifier to determine a classification of the received content, the DNN classifier including a clustering classification layer that clusters based on a latent feature vector representation of the content, the classification corresponding to one or more clusters that are closest to the latent feature vector providing the classification and a corresponding confidence.

Abstract: Techniques for multi-factor pathfinding are disclosed. A system receives multiple data sets that represent multiple physical properties associated with multiple three-dimensional coordinates of a physical environment. The system generates multiple weighted data layers corresponding to the physical properties, at least by applying weights to the data sets. The system generates a multi-factor costmap for the three-dimensional environment, based at least on the weighted data layers. The system determines a path through the three-dimensional environment, based at least on the multi-factor costmap.

Abstract: A neural network apparatus includes processing circuitry and memory. The memory stores a plurality of images of a target. The processing circuitry is configured to: access, from the memory, a first image and an identification of a centroid pixel of the target within the first image; generate, based on a geometry of the target and the centroid pixel, a confidence map indicating, for each pixel in the first image, a confidence value that the pixel includes the target; train, using the plurality of images of the target, including the first image and the confidence map, an artificial neural network to identify the target in visual data; and provide an output representing the trained artificial neural network.

Abstract: A neural network apparatus includes processing circuitry and memory. The memory stores a plurality of images of a target. The processing circuitry is configured to: access, from the memory, a first image and an identification of a centroid pixel of the target within the first image; generate, based on a geometry of the target and the centroid pixel, a confidence map indicating, for each pixel in the first image, a confidence value that the pixel includes the target; train, using the plurality of images of the target, including the first image and the confidence map, an artificial neural network to identify the target in visual data; and provide an output representing the trained artificial neural network.

Abstract: An apparatus for a wireless power transfer (“WPT”) pad heat management system includes a ferrite structure positioned adjacent to a coil configured to wirelessly transfer power. The apparatus includes a plurality of heat spreaders positioned along a length of a component of the ferrite structure. Each of the plurality of heat spreaders is non-metallic. The apparatus includes a trough shaped to surround at least a portion of each of the plurality of heat spreaders, wherein the trough is non-metallic. The apparatus includes a phase change material (“PCM”) in the trough where at least a portion of the heat spreaders extend into the PCM. The ferrite structure, coil, plurality of heat spreaders, trough and PCM are encased in a solid material, and each of the plurality of heat spreaders comprises a material that transfers heat from the component of the ferrite structure to the PCM.

Abstract: Infrared sensing systems having improved vibration cancelation, and methods of achieving improved vibration cancelation. In one example, an infrared sensing system includes an infrared sensor configured to produce a sensor output signal representative of a response of the infrared sensor to infrared excitation and vibration excitation, an accelerometer configured to provide an acceleration signal responsive to the vibration excitation, and a controller, including an adaptive digital filter, coupled to the infrared sensor and to the accelerometer, and configured to receive the acceleration signal and to adjust coefficients of the adaptive digital filter so as to minimize coherence between a residual signal and the acceleration signal, the residual signal being a difference between the sensor output signal and a filter output signal from the adaptive digital filter.

Abstract: Subject matter regards improving machine learning techniques using informed pseudolabels. A method can include receiving previously assigned labels indicating an expected classification for data, the labels having a specified uncertainty, generating respective pseudolabels for the data based on the previously assigned labels, the data, a class vector determined by an ML model, and a noise model indicating, based on the specified uncertainty, a likelihood of the previously assigned label given the class, and substituting the pseudolabels for the previously assigned labels in a next epoch of training the ML model.

Abstract: A rotary blade includes a blade tip assembly with a first bus bar segment and a main blade assembly with second bus bar segment. The first bus bar segment is connected to the blade tip assembly and the second bus bar segment is connected to the main blade assembly. The blade tip assembly is connected to the main blade assembly such that the first and second bus bar segments are longitudinally offset from one another. A low-profile interconnect spans the first and second bus bar segments for resistively generating less heat than the bus bar segments for a predetermined current flow.

Abstract: An aircraft heated floor panel includes a first face sheet, a second face sheet opposite the first face sheet, and core with an electrically conductive core portion. The electrically conductive core portion supports the first face sheet and the second face sheet, and is electrically insulated from the external environment to receive electrical power, resistively generate heat, and communicate heat to the first face sheet.

Abstract: Optical sensing systems having improved vibration cancelation, and methods of achieving improved vibration cancelation. In one example, an optical sensing system includes an optical sensor configured to produce an unprocessed sensor output signal representative of a response of the optical sensor to at least an optical signature of interest and a local vibration excitation, a reference sensor configured to provide a reference signal responsive to the local vibration excitation, and a controller, including an adaptive digital filter, coupled to the optical sensor and to the reference sensor, and configured to receive the reference signal and to adjust one or more coefficients of the adaptive digital filter to minimize coherence between a residual signal and the reference signal, the residual signal being a difference between the sensor output signal and a filter output signal from the adaptive digital filter.

Abstract: Infrared sensing systems having improved vibration cancelation, and methods of achieving improved vibration cancelation. In one example, an infrared sensing system includes an infrared sensor configured to produce a sensor output signal representative of a response of the infrared sensor to infrared excitation and vibration excitation, an accelerometer configured to provide an acceleration signal responsive to the vibration excitation, and a controller, including an adaptive digital filter, coupled to the infrared sensor and to the accelerometer, and configured to receive the acceleration signal and to adjust coefficients of the adaptive digital filter so as to minimize coherence between a residual signal and the acceleration signal, the residual signal being a difference between the sensor output signal and a filter output signal from the adaptive digital filter.

Abstract: A rotary blade includes a blade tip assembly with a first bus bar segment and a main blade assembly with second bus bar segment. The first bus bar segment is connected to the blade tip assembly and the second bus bar segment is connected to the main blade assembly. The blade tip assembly is connected to the main blade assembly such that the first and second bus bar segments are longitudinally offset from one another. A low-profile interconnect spans the first and second bus bar segments for resistively generating less heat than the bus bar segments for a predetermined current flow.

Abstract: An ice protection system (40) comprising electrothermal devices (41-44) having zones (41P-44P, 41L-44L, 41S-44S, 41T-44T) associated with respective surface regions of rotor blades. Each zone involves electric load (R) which converts supplied electrical power from one phase of a three-phase power system into heat. The combined load of the A-phase zones, the combined load of the B-phase zones, and the combined load of the C-phase zones are approximately equal.

Abstract: A method begins with a payment entity device receiving an accounts payable data file from a client entity processing module. The method continues with the payment entity device determining a method of payment for an account payable of the accounts payable data file in accordance with a payables profile. The method continues with the payment entity device initiating a first type of payment for the account payable when the method of payment is a first type; initiating a second type of payment for the account payable, when the method of payment is a second type, or initiating a third type of payment for the account payable when the method of payment is a third type.

Abstract: A method begins with a payment entity device receiving payables process data from a client device. The payables process data includes a list of a plurality of creditors and associated payment data, wherein, for a creditor of the plurality of creditors, the associated payment data includes at least one payment scheme for paying at least a portion of debt owed to the creditor via at least one of: a client credit card, a funds transfer, commercial paper, tangible consideration, and a debit account. The method continues with the payment entity device determining level of service for a client associated with the client device. The method continues with the payment entity device generating a payables profile for the client based on the payables process data in accordance with the level of service.

Abstract: A method begins by storing payment request data, which, for a client, includes a payables profile and an accounts payable data file. The method continues by receiving payment remittance information subsequent to initiation of a payment in accordance with the payment request data. The method continues by storing the payment remittance information. The method continues by consolidating the payment remittance information and the payment request data to produce consolidated payment data. The method continues by generating at least one report based on the consolidated payment data.