Abstract: Detecting user contact with one or more electrodes of a physiological signal sensor can be used to ensure physiological signals measured by the physiological signal sensor meet waveform characteristics (e.g., of a clinically accurate physiological signal). In some examples, a mobile and/or wearable device can comprise sensing circuitry, stimulation circuitry, and processing circuitry. The stimulation circuit can drive one or more stimulation signals on one or more electrodes, the resulting signal(s) can be measured (e.g., by the sensing circuitry), and the processing circuitry can determine whether a user is in contact with the electrode(s). Additionally or alternatively, in some examples, mobile and/or wearable device can comprise saturation detection circuitry, and the processing circuitry can determine whether the sensing circuitry is saturated.

Abstract: Systems and methods for enhancing text in images based on super-resolution are disclosed. A low resolution image is generated based on a high resolution image. A super resolution image is generated based on the low resolution image, using a super resolution model with a set of parameters. Based on the high resolution image and the super resolution image, a total loss function is computed based on: the set of parameters, a detection loss function, a recognition loss function, and a gradient loss function of the high resolution image and the super resolution image. A trained super resolution model is generated with an optimized set of the parameters that minimizes the total loss function. Text in at least one image is enhanced using the trained super resolution model.

Abstract: A rotor assembly includes a rotor core having a rotatable shaft and defining at least one rotor post, a winding wound around the post that defines an end turn, and a winding support assembly including a winding support assembly coupled to the rotatable shaft. The end turn defines a set of channels therethrough and is disposed in a first chamber defined by the winding support assembly.

Abstract: A method of cooling a set of rotor winding end turns of a rotor assembly can include directing a fluid coolant flow to a coil support assembly, delivering the fluid coolant flow, by a first coolant distribution ring of the coil support assembly, radially outward toward rotor winding end turns; and delivering the fluid coolant flow, by a coil support disc of the coil support assembly, axially outward toward the rotor winding end turns. The method can further include expelling fluid coolant flow by a second coolant distribution ring radially outward toward a set of stator windings, and expelling, by the coil support disc, the fluid coolant flow axially outward toward the rotor core.

Abstract: Segmentation first breaks the images into segments or regions, with the segments of the region having text or symbols. The segmented image is separately applied to two different CNN-based models. Each model produces text boxes where potential text might exist. Then, a selective NMS algorithm is applied to the output of each model to produce a final group of text regions. These text regions are analyzed and actions are taken.

Abstract: A rotor assembly comprises a rotor core comprising a first end and an opposing second end, and an outwardly facing peripheral surface having a set of rotor teeth projecting outwardly therefrom to define a set of rotor slots therebetween. Each rotor tooth can comprise a first rotor tooth wall, an opposing second rotor tooth wall, and a distal tip between the first rotor tooth wall and second rotor tooth wall. Each respective rotor slot can be defined by a respective first rotor tooth wall of a first rotor tooth, and a respective second rotor tooth wall of a second rotor tooth, and includes a first rotor core conduit extending axially along a length of the first rotor tooth. The first rotor core conduit can comprise a first rotor core conduit inlet at the rotor core first end and a first rotor core conduit inlet at the rotor core second end, the rotor core conduit defining a first taper that is narrower at first rotor core conduit inlet than the first rotor core conduit outlet.

Abstract: This document describes techniques and apparatuses for automatic white-balance for a camera system. The techniques and apparatuses utilize a precursor image to detect one or more detected faces and determine a tone. The camera system retrieves tonal data based on a group of images determined to contain a same face as the detected face. Based on this tonal data, a difference in white balance is determined based on the difference in tone of the detected face within the precursor image and the associated tonal data. Camera settings are adjusted based on the difference in white balance to enable capture of an image having an improved tone.

Abstract: Embodiments include a system for analyzing breathing parameters of a user. The system can include an optical sensing unit configured to detect movement of a torso of the user and output one or more signals indicative of the movement of the torso. The system can also include an electronic device configured to request for the user to breathe according to a breathing profile during a time period. The system can include a processing unit configured to receive the one or more signals generated by the optical sensing unit during the time period and determine an adherence metric using the one or more signals and the breathing profile, where the adherence metric indicating a correspondence between movement of the torso and the breathing profile requested during the time period.

Abstract: A method can include determining a training dataset for a first machine learning module based on historical change data by detecting and correcting correctable label noises in the historical change data and after detecting and correcting the correctable label noises, determining the training dataset based on a label-error-free portion of the historical change data. The method further can include training the first machine learning module based on the training dataset to determine a respective risk score associated with a respective change request. The method additionally can include determining, via the first machine learning module, as trained, a first risk score associated with a first change request. Further, the method can include determining a change approval based on the first risk score and a risk threshold. Then, the method can include transmitting the change approval to cause an implementation of the first change request. Other embodiments are disclosed.

Abstract: Systems and methods for enhancing text in images based on super-resolution are disclosed. A low resolution image is generated based on a high resolution image. A super resolution image is generated based on the low resolution image, using a super resolution model with a set of parameters. Based on the high resolution image and the super resolution image, a total loss function is computed based on: the set of parameters, a detection loss function, a recognition loss function, and a gradient loss function of the high resolution image and the super resolution image. A trained super resolution model is generated with an optimized set of the parameters that minimizes the total loss function. Text in at least one image is enhanced using the trained super resolution model.

Abstract: Embodiments are directed to a wearable device that can include a housing, a touch-sensitive display coupled to the housing, a band coupled to the housing and configured to secure the housing to a user, and a plurality of electrodes that contact the user when the wearable device is worn by the user. The wearable device can be configured to cause the plurality of electrodes to apply electrical signals to the user, measure the electrical signals from the user using the plurality of electrodes, generate a set of impedance data from the measured electrical signals, and identify one or more respiratory cycles using the set of impedance data.

Abstract: Detecting user contact with one or more electrodes of a physiological signal sensor can be used to ensure physiological signals measured by the physiological signal sensor meet waveform characteristics (e.g., of a clinically accurate physiological signal). In some examples, a mobile and/or wearable device can comprise sensing circuitry, stimulation circuitry, and processing circuitry. The stimulation circuit can drive one or more stimulation signals on one or more electrodes, the resulting signal(s) can be measured (e.g., by the sensing circuitry), and the processing circuitry can determine whether a user is in contact with the electrode(s). Additionally or alternatively, in some examples, mobile and/or wearable device can comprise saturation detection circuitry, and the processing circuitry can determine whether the sensing circuitry is saturated.

Abstract: A rotor assembly includes a rotor core having a rotatable shaft and defining at least one rotor post, a winding wound around the post that defines an end turn, and a coil support assembly including a coil support disc rotatably coupled to the rotatable shaft and comprising a radially inner portion and a radially outer portion coupled via a set of spokes. The radially outer portion overlies an end turn and defines a set of channels therethrough sized and disposed to receive a flow of fluid from the set of rotor windings.

Abstract: Segmentation first breaks the images into segments or regions, with the segments of the region having text or symbols. The segmented image is separately applied to two different CNN-based models. Each model produces text boxes where potential text might exist. Then, a selective NMS algorithm is applied to the output of each model to produce a final group of text regions. These text regions are analyzed and actions are taken.

Abstract: This application relates to systems and methods for automatically correcting labels in untrusted data based on a small sample of trusted data in a training database. In some examples, training data may be divided into a trusted dataset and an untrusted dataset using stratified sampling. An adversarial algorithm may be used to reassign labels in the data samples associated with the untrusted data based on a set of features in the data and labels in the trusted dataset. The untrusted dataset with the reassigned labels may then be used to train a machine learning model.

Abstract: Segmentation first breaks the images into segments or regions, with the segments of the region having text or symbols. The segmented image is separately applied to two different CNN-based models. Each model produces text boxes where potential text might exist. Then, a selective NMS algorithm is applied to the output of each model to produce a final group of text regions. These text regions are analyzed and actions are taken.

Abstract: A method of cooling rotor winding end turns extending from a rotor core includes coupling a collar to a rotatable shaft of the rotor core. The collar has a first wall confronting the rotor core, a second wall spaced from and opposing the first wall, a third wall between the first and second walls defining a set of apertures therethrough, a first cavity cooperatively defined by the first, second and third walls, having a first opening opposing the third wall, a fourth wall circumscribing the third wall, a second cavity cooperatively defined by the second, third and fourth walls, the second cavity defining a second opening opposing the second wall. The first and second cavities are in fluid communication, and the rotor winding end turns are receivable into the second cavity. The method includes directing a coolant into the first cavity, and delivering the coolant to the second cavity.

Abstract: Embodiments include a system for measuring respiratory function of a user. The system can include an optical sensing unit configured to detect movement of a torso of the user. The system can include an electronic device configured to provide a first request for the user to breathe at a first rate during a first time period and a second request for the user to breathe at a second rate during a second time period. The system can include a processing unit configured to determine a first respiration parameter based on the movement of the torso during the first time period and determine a second respiration parameter based on the movement of the torso during the second time period. The processing unit can determine a level of respiratory function based on the first respiration parameter and the second respiration parameter.

Abstract: A rotor assembly includes a rotor core having a rotatable shaft and defining at least one rotor post, a winding wound around the post that defines an end turn, and a coil support assembly including a coil support disc rotatably coupled to the rotatable shaft and comprising a radially inner portion and a radially outer portion coupled via a set of spokes. The radially outer portion overlies an end turn and defines a set of channels therethrough sized and disposed to receive a flow of fluid from the set of rotor windings.

Abstract: A rotor assembly comprises a rotor core comprising a first end and an opposing second end, and an outwardly facing peripheral surface having a set of rotor teeth projecting outwardly therefrom to define a set of rotor slots therebetween. Each rotor tooth can comprise a first rotor tooth wall, an opposing second rotor tooth wall, and a distal tip between the first rotor tooth wall and second rotor tooth wall. Each respective rotor slot can be defined by a respective first rotor tooth wall of a first rotor tooth, and a respective second rotor tooth wall of a second rotor tooth, and includes a first rotor core conduit extending axially along a length of the first rotor tooth. The first rotor core conduit can comprise a first rotor core conduit inlet at the rotor core first end and a first rotor core conduit inlet at the rotor core second end, the rotor core conduit defining a first taper that is narrower at first rotor core conduit inlet than the first rotor core conduit outlet.