Abstract: A method for training a speech recognition model includes obtaining sample utterances of synthesized speech in a target domain, obtaining transcribed utterances of non-synthetic speech in the target domain, and pre-training the speech recognition model on the sample utterances of synthesized speech in the target domain to attain an initial state for warm-start training. After pre-training the speech recognition model, the method also includes warm-start training the speech recognition model on the transcribed utterances of non-synthetic speech in the target domain to teach the speech recognition model to learn to recognize real/human speech in the target domain.

Abstract: A weight distribution measurement system includes a portable measuring device configured to measure a linear distance and store a plurality of measurements. The portable measuring device is also configured to measure a first distance between a first location on a vehicle supported by an axle of the vehicle and a second location, wherein the first distance is measured prior to loading the vehicle. The portable measuring device is also configured to measure a second distance between the first location and the second location, wherein the second distance is measured after loading a vehicle. And the portable measuring device can measure a third distance between the first location and the second location, wherein the third distance is measured after loading a vehicle and after a weight distribution mechanism has been engaged, and then calculate the percent weight distribution over the axles of the vehicle using the portable measuring device and display the percent weight distribution.

Abstract: Embodiments relate to binaural spatialization of more than two sound sources on two audio channels of an audio system. Sound signals each emitted from a corresponding sound source are collected, and a respective virtual position within an angular range of a sound scene is assigned to each sound source. Multi-source audio signals are generated by panning each sound signal according to the respective virtual position. A first multi-source audio signal is spatialized to a first direction to generate a first left signal and a first right signal. A second multi-source audio signal is spatialized to a second direction to generate a second left signal and a second right signal. A binaural signal is generated using the first left signal, the second left signal, the first right signal, and the second right signal. The binaural signal is such that each sound source appears to originate from its respective virtual position.

Abstract: A dynamic control valve assembly for use in filling a liquid carbon dioxide storage and gas delivery system is provided, the assembly comprising: a valve body; an end nut with an inlet port for receiving liquid carbon dioxide; a chamber; an inlet cavity; a liquid port; a gas port; and a dynamic compound valve stem assembly for blocking the gas port while liquid carbon dioxide is delivered through the inlet port and allowing the liquid carbon dioxide to flow through the liquid port for storage in a liquid cylinder, and open the gas port and block the inlet port in order to allow carbon dioxide gasses from boiling liquid carbon dioxide within the liquid cylinder to pass through the gas port for storage in a gas cylinder until system pressure and temperature equilibrium is reached. The dynamic compound valve stem assembly comprises: a stem body having an inlet port poppet and a gas port poppet; an inlet cavity collar; and in some embodiments a collar biasing spring.

Abstract: A method is provided for displaying an augmented reality (AR) representation of physical effects and property damage resulting from a parametric earthquake event. The method includes scanning, using one or more sensors of a user device, a scene in proximity to a user. The method also includes identifying a background and objects in the scene. The method further includes creating an AR background for the background of the scene and AR objects for the objects in the scene. In addition, the method includes receiving at least one seismic characteristic from the user through the display of the user device. The method additionally includes displaying at least one seismic effect on the AR objects and the AR background in the scene based on the at least one received seismic characteristic.

Abstract: A process for manufacturing glazed ceramic tiles from a substrate of green clay having at least one groove extending within the surface of the tile, wherein the groove is imparted to the tile while the substrate is still green clay before the substrate is bisque or glaze fired. A tile fabricated using this process is also part of the invention.

Abstract: Catalyst complexes include a zinc hexacyanocobaltate with M5 metal and M6 metal or semi-metal phases, wherein M5 metal is gallium, hafnium, manganese, titanium and/or indium and the M6 metal or semi-metal is one or more of aluminum, magnesium, manganese, scandium, molybdenum, cobalt, tungsten, iron, vanadium, tin, titanium, silicon and zinc and is different from the M5 metal. The catalysts are highly efficient propylene oxide polymerization catalysts characterized by rapid activation, short times to the onset of rapid polymerization and high polymerization rates once rapid polymerization has begun.

Abstract: A hydraulic tool includes a rotor rotatably disposed within a stator. At least an inner portion of the stator and/or at least an outer portion of the rotor is configured to be installed in a drill string in either of two inverted orientations along a longitudinal axis of the hydraulic tool. The rotor is configured to rotate within the stator in either of the two orientations. A method includes disposing a rotor within a cavity defined by a stator, passing a fluid through the stator to rotate the rotor, and reversing the stator or the rotor. A drilling system includes a fluid source, a hydraulic tool, a drive shaft operatively associated with the rotor of the hydraulic tool, and a drill bit operatively associated with the drive shaft.

Abstract: An exemplary system, method, and computer-accessible medium can include receiving first information related to a first transaction associated with a financial account, determining a first network coverage area associated with the first transaction, receiving second information related to a second transaction associated with the financial account, determining a second network coverage area associated with the second transaction, flagging the second transaction as potentially fraudulent based on the first network coverage area and the second network coverage area. The financial account can be a credit card account. The second transaction can be flagged if the first network coverage area and the second network coverage area do not overlap. The first network coverage area can be based on a coverage area for a first satellite and the second network coverage area can be based on the coverage area for a second satellite, where the first satellite can be different than the second satellite.

Abstract: Generally, the present disclosure is directed to systems and methods that train machine-learned models (e.g., artificial neural networks) to perform perceptual or cognitive task(s) based on biometric data (e.g., brain wave recordings) collected from living organism(s) while the living organism(s) are performing the perceptual or cognitive task(s). In particular, aspects of the present disclosure are directed to a new supervision paradigm, by which machine-learned feature extraction models are trained using example stimuli paired with companion biometric data such as neural activity recordings (e g electroencephalogram data, electrocorticography data, functional near-infrared spectroscopy, and/or magnetoencephalography data) collected from a living organism (e.g., human being) while the organism perceived those examples (e.g., viewing the image, listening to the speech, etc.).

Abstract: A light inspection system positions an autonomous vehicle (AV) in a field of view of a camera such that the camera captures an image of a light of the AV. The light inspection system instructs a camera to capture an image of the light while the light is switched on. The light inspection system receives the captured image, determines a luminance of the light based on the image, and determines to service the light in response to the luminance of the light being below a threshold luminance.

Abstract: A system includes a vent housing configured to be installed on a lower housing of a battery module at a first side of the vent housing. The vent housing has a main body having an opening on a second side of the vent housing and an internal chamber coupled to the opening. The internal chamber includes a first wall having an internal burst vent configured to open at a first pressure threshold and a second wall having a ventilation vent comprising a gas-selective permeability layer.

Abstract: A foldable toilet footrest that includes a platform with a first leg and a second leg coupled thereto by hinges. The hinges can include a releasable locking system that allows the first leg and the second leg by to be fixed in an unfolded, use position or a folded, non-use position.

Abstract: A garment for assisting a user holding an infant is provided. The garment may include a first portion, a second portion, a first sleeve, and a second sleeve. The second portion may be configured to overlap the first portion when the garment is worn by the user. The first sleeve and the second sleeve may each extend from the first portion and the second portion, respectively. The first portion may define an elongated opening. The elongated opening may be positioned to provide access through the opening and to the infant from the outside of the garment while the first portion remains covered by the second portion. The garment may also include at least one fastener attached thereto. The at least one fastener may be positioned relative to the opening and may comprise at least one loop to secure at least part of a tubing or wire extending from the infant. The garment may also comprise a locking mechanism that can adjust the loop to the size of the tubing.

Abstract: An exemplary system can generate an alert to a caregiver or other concerned parties when a usage pattern of a multimedia device differs by a preset amount. The multimedia device can record typical usage of the device and determine a pattern. The pattern is compared with real-time usage to determine when a different pattern is being done and if it is outside a preset amount generate an alert to a caregiver. This allows for non-intrusive monitoring of individuals that use the multimedia device on a regular basis.

Abstract: There is included a method and apparatus comprising computer code configured to cause a processor or processors to perform obtaining video data, determining an access unit delimiter syntax of at least one network abstraction layer (NAL) unit of the video data, determining whether the access unit delimiter syntax indicates a slice type value of the NAL unit, determining whether the access unit delimiter syntax of the NAL unit indicates at least one of a video coding layer (VCL) value and an access unit order count value of the NAL unit, and signaling access unit boundary values of the NAL unit according to whether the access unit delimiter syntax indicates at least one of the slice type value, the VCL value, and the access unit order count value.

Abstract: A system for a power take-off mechanism for a powertrain system is provided. The system includes an electrically powered torque generating device including a torque generating device output shaft and a transmission output shaft receiving mechanical power from the torque generating device output shaft. The system further includes a clutch selectively disengaging the transmission output shaft from the torque generating device output shaft and a power take-off module receiving mechanical power from the torque generating device.

Abstract: A lubricating fluid comprising a base oil, which utilizes carboxylic esters of carboxyl di-end-capped-polyethylene glycols, or a mixture thereof with carboxylic esters of carboxyl di-end-capped-polytetramethylene glycols, to improve low temperature performance, as well as minimize elastohydrodynamic shear strength, and enables the production of high efficiency fluids with enhanced low temperature properties for machines or machine elements that operate in the elastohydrodynamic regime in lubrication.

Abstract: The execution-time reporting of telemetry of execution of a software program. Subscribers submit subscriptions to telemetry of the software program. As each subscription is received, the telemetry scope of the subscription is evaluated to determine what portion of an object model is to be augmented. The augmented portion will include portion(s) related to the scope of telemetry subscribed to in the subscription. Thereafter, that portion of the object model is indeed augmented as execution of the computer program proceeds further. Subsequently, telemetry reports are generated for a subscription based on the interpretation of a defined subscription-specific portion of the object model.

Abstract: A method for switching a configuration of an enhanced reality headset is provided. The method includes identifying a scenario to switch a configuration of an enhanced reality headset between a virtual reality configuration, an augmented reality configuration, and a direct reality configuration, wherein the virtual reality configuration and the augmented reality configuration include a computer generated image, and the direct reality configuration and the augmented reality configuration include a real-time image of an environment of a user of the enhanced reality headset. The method includes providing a notification to a user of the enhanced reality headset including an intent to switch the configuration of the enhanced reality headset, and switching the configuration of the enhanced reality headset according to the scenario.