Abstract: A device having a reference transistor fabricated within the same semiconductor substrate as a primary transistor (e.g., configured for use in a radiofrequency amplifier or other active circuit) has a shared metallization area coupled to a current terminal of both transistors configured to shield a control terminal of the reference transistor from coupling of alternating current interference from alternating currents within the primary transistor.

Abstract: Composite airfoils and methods for forming composite airfoils are provided. For example, a composite airfoil of a gas turbine engine comprises opposite pressure and suction sides extending radially along a span from a root to a tip, which define opposite radial extremities of the airfoil. The composite airfoil further comprises a body section and a tip section, which includes the tip, that each extend radially along the span. The composite airfoil is formed from a composite material comprising fibers disposed in a matrix material. The tip section has a tip fiber volume, and the body section has a body fiber volume that is greater than the tip fiber volume. Another composite airfoil comprises a tip cap applied over the tip that tapers from a first end to a second end such that each of the pressure and suction side walls of the tip cap narrows from a first thickness to a second thickness.

Abstract: A pre-assembled decking panel and strut assembly for data center ceilings and roofs including a decking panel that is a sheet metal in a shape of a wave and an elongated strut member configured such that hanging systems can be attached to the elongated strut member. The peaks and valleys of the decking panel and the strut member extend in a first direction and are spaced in parallel in a second direction perpendicular to the first direction. A top wall of the strut member is affixed to a peak of the decking panel.

Abstract: A forceps is provided and includes a housing having a shaft. An end effector assembly operatively connects to a distal end of the shaft and includes a pair of first and second jaw members. One or both of the first and second jaw members is movable relative to the other jaw member from a clamping position to an open position. A resilient member operably couples to at least one of the first and second jaw members. The resilient member is configured to bias the first and second jaw members in the clamping position and provide a closure force on tissue disposed therebetween.

Abstract: Disclosed are chimeric polypeptides based on viral membrane fusion proteins. More particularly, the present invention discloses chimeric polypeptides that comprise a virion surface exposed portion of a viral fusion protein and a heterologous structure-stabilizing moiety, and to complexes of those chimeric polypeptides. The present invention also discloses the use of these complexes in compositions and methods for eliciting an immune response to a fusion protein of an enveloped virus, or complex of the fusion protein, and/or for treating or preventing an enveloped virus infection. The present invention further discloses the use of the heterologous structure-stabilizing moiety for oligomerizing heterologous molecules of interest.

Abstract: Data storage devices configured to exploit generative-adversarial-networks (GANs) are described herein, including super-resolution GANs (SRGANs). In some examples, a GAN-based decoding (reconstruction) procedure is implemented within a data storage controller to replace or supplement an error correction coding (ECC) decoding procedure to permit a reduction in the number of parity bits used while storing the data. In other examples, soft bit information is exploited using GANs during decoding. A dissimilarity matrix may be generated to represent differences between an initial image and a GAN-reconstructed image, with matrix values mapped into low-density parity check (LDPC) codewords to facilitate LDPC decoding of data. In still other examples, confidence information obtained from a GAN is incorporated into image pixels. In some examples, GAN reconstruction of data is limited to modifying valley bits. Multiple GANs may be used in parallel with their outcome aggregated. System and method examples are provided.

Abstract: Data storage devices configured to exploit generative-adversarial-networks (GANs) are described herein, including super-resolution GANs (SRGANs). In some examples, a GAN-based decoding (reconstruction) procedure is implemented within a data storage controller to replace or supplement an error correction coding (ECC) decoding procedure to permit a reduction in the number of parity bits used while storing the data. In other examples, soft bit information is exploited using GANs during decoding. A dissimilarity matrix may be generated to represent differences between an initial image and a GAN-reconstructed image, with matrix values mapped into low-density parity check (LDPC) codewords to facilitate LDPC decoding of data. In still other examples, confidence information obtained from a GAN is incorporated into image pixels. In some examples, GAN reconstruction of data is limited to modifying valley bits. Multiple GANs may be used in parallel with their outcome aggregated. System and method examples are provided.

Abstract: Data storage devices configured to exploit generative-adversarial-networks (GANs) are described herein, including super-resolution GANs (SRGANs). In some examples, a GAN-based decoding (reconstruction) procedure is implemented within a data storage controller to replace or supplement an error correction coding (ECC) decoding procedure to permit a reduction in the number of parity bits used while storing the data. In other examples, soft bit information is exploited using GANs during decoding. A dissimilarity matrix may be generated to represent differences between an initial image and a GAN-reconstructed image, with matrix values mapped into low-density parity check (LDPC) codewords to facilitate LDPC decoding of data. In still other examples, confidence information obtained from a GAN is incorporated into image pixels. In some examples, GAN reconstruction of image data is limited to modifying valley bits. Multiple GANs may be used in parallel with their outcome aggregated.

Abstract: A system includes an energy storage system having a plurality of batteries and a monitoring system operably coupled to the energy storage system. The monitoring system is configured to receive a first measurement corresponding to a characteristic from at least one of the plurality of batteries, determine a baseline from the first measurement, receive a second measurement from the at least one of the plurality of batteries, and compare the second measurement with the baseline to identify an issue with the at least one of the plurality of batteries. The monitoring system is also configured to perform a remediation action in response to the identified issue.

Abstract: A method for producing hydrogen product having a low carbon intensity is provided. The method includes the steps of: (a) converting a hydrocarbon feedstock to a hydrogen product using a hydrocarbon reforming process; (b) providing at least some of the required energy for the hydrogen production process from a biomass power plant; and (c) processing one or more flue gas streams from the biomass power plant in a carbon capture unit to reduce CO2e emissions. The hydrogen product has a carbon intensity preferably less than about 1.0 kg CO2e/kg H2, more preferably less than 0.45 kg CO2e/kg H2, and most preferably less than 0.0 kg CO2e/kg H2.

Abstract: An end-to-end automatic speech recognition (ASR) system can be constructed by fusing a first ASR model with a transformer. The input of the transformer is a learned layer generated by the first ASR model. The fused ASR model and transformer can be treated as a single end-to-end model and trained as a single model. In some embodiments, the end-to-end speech recognition system can be trained using a teacher-student training technique by selectively truncating portions of the first ASR model and/or the transformer components and selectively freezing various layers during the training passes.

Abstract: In one embodiment of an audio system, a transducer can be coupled to a passive acoustic directional amplifier to provide various benefits and improvements, including improvements to: speech intelligibility, signal-to-noise ratio, effective equivalent input noise, at-a-distance acoustic signal reception, and directional preference. In another embodiment, the shape of an interior surface of a passive acoustic directional amplifier is provided. In another embodiment, the material properties of an interior surface of a passive acoustic directional amplifier are provided.

Abstract: A method includes determining, by a sensor node in communication with a communication network, an evacuation condition, retrieving by a decision node, occupancy information pertaining to an area, retrieving area data, and determining an evacuation route, at least in part using the evacuation condition, occupancy information, and area data, and communicating the determined evacuation route to area occupants.

Abstract: Self-closing safety gate embodiments are disclosed herein. The self-closing safety gates include a hinge plate configured to rotate about a vertical support member and engage a hoop portion. In some embodiments, the hinge plate includes at least one support arm portion with a channel. In such embodiments, the channel can be configured to receive and frictionally engage a hoop arm of the hoop portion with a compressive fastener. In some embodiments, the self-closing safety gate can be damped.

Abstract: A display system can include a head-mounted display configured to project light to an eye of a user to display virtual image content at different amounts of divergence and collimation. The display system can include an inward-facing imaging system that images the user's eye and processing electronics that are in communication with the inward-facing imaging system and that are configured to obtain an estimate of a center of rotation of the user's eye. The display system may render virtual image content with a render camera positioned at a determined position relative to the determined position of the center of rotation of said eye.

Abstract: A deployment device for positioning a soft tissue repair prosthesis includes a self-expanding support body releasably attachable to the prosthesis. The support body may be removably insertable into a pocket of the prosthesis. A handle coupled to the support body facilitates positioning the patch and/or removal of the support body from the pocket. The support body may substantially occupy the pocket in an expanded configuration. The handle may be arranged to direct a pulling force to the outer peripheral edge of the support body and/or cause a portion thereof to be pulled downward and below the body during withdrawal of the deployment device from the prosthesis. The support body may include support segments pivotally coupled together and foldable to collapse the support body for insertion into and removal from the pocket. One or more resilient support members may be provided for collapsing and expanding the support body and the prosthesis.

Abstract: A gas turbine engine is provided. The gas turbine engine includes a turbomachine defining an engine inlet to an inlet duct, a fan duct inlet to a fan duct, and a core inlet to a core duct; a primary fan driven by the turbomachine; and a secondary fan located downstream of the primary fan within the inlet duct. The gas turbine engine defines a thrust to power airflow ratio between 3.5 and 100 and a core bypass ratio between 0.1 and 10, wherein the thrust to power airflow ratio is a ratio of an airflow through a bypass passage over the turbomachine plus an airflow through the fan duct to an airflow through the core duct, and wherein the core bypass ratio is a ratio of the airflow through the fan duct to the airflow through the core duct.

Abstract: There is a region of interest of a synthetic image depicting an object from a class of objects. A trained neural image generator, having been trained to map embeddings from a latent space to photorealistic images of objects in the class, is accessed. A first embedding is computed from the latent space, the first embedding corresponding to an image which is similar to the region of interest while maintaining photorealistic appearance. A second embedding is computed from the latent space, the second embedding corresponding to an image which matches the synthetic image. Blending of the first embedding and the second embedding is done to form a blended embedding. At least one output image is generated from the blended embedding, the output image being more photorealistic than the synthetic image.