Abstract: Debiasing pre-trained sentence encoders with probabilistic dropouts may be performed by various systems, services, or applications. A sentence may be received, where the words of the sentence may be provided as tokens to an encoder of a machine learning model. A token-wise correlation using semantic orientation may be determined to determine a bias score for the tokens in the input sentence. A probability of dropout that for tokens in the input sentence may be determined from the bias scores. The machine learning model may be trained or tuned based on the probabilities of dropout for the tokens in the input sentence.

Abstract: A method and system for detecting pulsed communication signals, the method and system includes obtaining a time-frequency representation of a received training data signal, the received training data signal representative of a known/training bitstream associated with a transmitted pulse signal; using a scalogram process to convert the time-frequency representation of the received training data signal into a plurality of training images, and training a deep learning architecture platform with the plurality of training images to generate a classification model representative of a plurality of high bit states and a plurality of low bit states included in the plurality of training images.

Abstract: Debiasing pre-trained sentence encoders with probabilistic dropouts may be performed by various systems, services, or applications. A sentence may be received, where the words of the sentence may be provided as tokens to an encoder of a machine learning model. A token-wise correlation using semantic orientation may be determined to determine a bias score for the tokens in the input sentence. A probability of dropout that for tokens in the input sentence may be determined from the bias scores. The machine learning model may be trained or tuned based on the probabilities of dropout for the tokens in the input sentence.

Abstract: In one aspect, a sleeper bunk is provided that includes a support for a mattress. The support includes an assembly of support portions each having at least one surface arranged to support an underside of the mattress. The sleeper bunk includes at least one mount connected to the support and configured to facilitate movement of the support and the mattress thereon between a raised orientation and a lowered orientation. The support portions have overlapping portions configured to distribute a load from the mattress between the support portions.

Abstract: In one embodiment, a vehicle wheel cover assembly is provided that includes a mount having an attachment portion with a release configuration that permits the attachment portion to engage inboard and outboard surfaces of a wheel hub flange of a wheel and a secured configuration that fixes the attachment portion to the wheel hub flange. The wheel cover assembly includes a retainer shiftable relative to the attachment portion of the mount from an unlocked position wherein the retainer permits the attachment portion to be reconfigured between the release configuration and the secured configured configuration to a locked position wherein the retainer maintains the attachment portion in the secured configuration. The wheel cover assembly further includes a cover assembly configured to be releasably secured to the mount to cover an opening of the wheel.

Abstract: Debiasing pre-trained sentence encoders with probabilistic dropouts may be performed by various systems, services, or applications. A sentence may be received, where the words of the sentence may be provided as tokens to an encoder of a machine learning model. A token-wise correlation using semantic orientation may be determined to determine a bias score for the tokens in the input sentence. A probability of dropout that for tokens in the input sentence may be determined from the bias scores. The machine learning model may be trained or tuned based on the probabilities of dropout for the tokens in the input sentence.

Abstract: In one aspect, a sleeper bunk is provided that includes rails having longitudinal lengths and a lateral spacing therebetween. The sleeper bunk has a plurality of support members connecting the rails. Each support member is secured to the rails and extends across the lateral spacing. The support members have a plurality of alternating lateral strengthening ribs and lateral channels. The lateral strengthening ribs of the support members are configured to support a mattress thereon. The sleeper bunk also includes a mount operable to facilitate movement of the rails and the support members between raised and lowered positions.

Abstract: In one aspect, a sleeper bunk is provided that includes a support for a mattress. The support includes an assembly of support portions each having at least one surface arranged to support an underside of the mattress. The sleeper bunk includes at least one mount connected to the support and configured to facilitate movement of the support and the mattress thereon between a raised orientation and a lowered orientation. The support portions have overlapping portions configured to distribute a load from the mattress between the support portions.

Abstract: In one embodiment, a vehicle wheel cover assembly is provided that includes a mount having an attachment portion with a release configuration that permits the attachment portion to engage inboard and outboard surfaces of a wheel hub flange of a wheel and a secured configuration that fixes the attachment portion to the wheel hub flange. The wheel cover assembly includes a retainer shiftable relative to the attachment portion of the mount from an unlocked position wherein the retainer permits the attachment portion to be reconfigured between the release configuration and the secured configured configuration to a locked position wherein the retainer maintains the attachment portion in the secured configuration. The wheel cover assembly further includes a cover assembly configured to be releasably secured to the mount to cover an opening of the wheel.

Abstract: A wheelchair anti-rollback device utilizes one-way roller clutches which incorporate a selectable free-rotation functionality such that the wheelchair user can select between anti-rollback and free-wheeling modes. The anti-rollback mode is of utility when the wheelchair user is propelling up ramps. The modal selection is by means of a knob positioned on the centerline rotation axis of the mechanism. The mechanism also included functionality that allows mode change from anti-rollback to free-wheeling to be made at any time, including while the chair is on an inclined ramp. The invention can incorporated into the hub of the rear wheels of a wheelchair or into the toggle mechanism of typically used to actuate parking brakes. When incorporated into a wheelchair, the invention gives the wheelchair user greater control and ease in ascending ramps.

Abstract: A system performs search and retrieval. The system monitors one or more user interface (“UI”) fields configured to receive text input in a UI. The system determines that the one or more UI fields are being used to enter a textual description, and performs a search on a knowledge base based on document similarity to identify documents that are similar to a portion of the textual description that has already been entered in the one or more UI fields. The system then provides one or more of the documents in a UI field of the UI, and repeats the monitoring, the determining, the performing, and the providing.

Abstract: Embodiments make novel use of random data structures to facilitate streaming inference for a Latent Dirichlet Allocation (LDA) model. Utilizing random data structures facilitates streaming inference by entirely avoiding the need for pre-computation, which is generally an obstacle to many current “streaming” variants of LDA as described above. Specifically, streaming inference—based on an inference algorithm such as Stochastic Cellular Automata (SCA), Gibbs sampling, and/or Stochastic Expectation Maximization (SEM)—is implemented using a count-min sketch to track sufficient statistics for the inference procedure. Use of a count-min sketch avoids the need to know the vocabulary size V a priori. Also, use of a count-min sketch directly enables feature hashing, which addresses the problem of effectively encoding words into indices without the need of pre-computation. Approximate counters are also used within the count-min sketch to avoid bit overflow issues with the counts in the sketch.

Abstract: The present invention relates generally to specially designed magnesium-aluminum (“Mg—Al”) alloys and medical uses thereof in a hemodynamic environment, and, more particularly to devices for sealing an opening formed through biological tissue (such as blood vessels) made from such specially designed magnesium-aluminum alloys.

Abstract: A ready-to-eat, shelf-stable tater-tot type snack food made from one or more of the following food products: vegetable (including starches from root vegetables), fruit, meat, bean, or meat analog. The tater tot-type snack food has a crispy outer layer or layers and crunchiness throughout the tater tot-type snack food. The tater tot-type snack food does not require cooking or warming prior to consumption. The tater tot-type snack food has extended shelf stability, especially when packaged in an oxygen free environment. The tater tot-type snack food can be manufactured using either freeze drying or vacuum frying techniques.

Abstract: A laser system for generating optical pulses at an operating wavelength of the laser system. The system has an optical resonator comprising first and second reflectors, and a tapered optical fiber disposed between the first and second reflectors. The tapered optical fiber has a core which has a tapered input section which tapers from single mode to multimode at the laser operating wavelength, an inner section of substantially constant diameter capable of supporting multiple modes at the laser operating wavelength and a tapered output section which tapers from a first diameter to a second diameter that is smaller than the first diameter.

Abstract: The present invention relates generally to specially designed magnesium-aluminum (“Mg—Al”) alloys and medical uses thereof in a hemodynamic environment, and, more particularly to devices for sealing an opening formed through biological tissue (such as blood vessels) made from such specially designed magnesium-aluminum alloys.

Abstract: A wheel retraction mechanism for use on travel luggage, the retraction mechanism causing the extension and retraction of one or more wheels upon relative movement of two adjacent members. The members comprise two engaging actuation surfaces which, upon relative longitudinal movement, create a movement in a different direction, which can be used to extend or retract wheels from a piece of travel luggage.

Abstract: A system performs search and retrieval. The system monitors one or more user interface (“UI”) fields configured to receive text input in a UI. The system determines that the one or more UI fields are being used to enter a textual description, and performs a search on a knowledge base based on document similarity to identify documents that are similar to a portion of the textual description that has already been entered in the one or more UI fields. The system then provides one or more of the documents in a UI field of the UI, and repeats the monitoring, the determining, the performing, and the providing.

Abstract: A wheel retraction mechanism for use on travel luggage, the retraction mechanism causing the extension and retraction of one or more wheels upon relative movement of two adjacent members. The members comprise two engaging actuation surfaces which, upon relative longitudinal movement, create a movement in a different direction, which can be used to extend or retract wheels from a piece of travel luggage.