Abstract: Systems and methods provides for dialog management in real-time rather than turn taking. An example method included generating first candidate responses to triggering event. The triggering event may be receipt of a live stream chunk for the dialog or receipt of a backend response to a previous backend request for a dialog shema. The method also includes updating a list of candidate responses that are accepted or pending with at least on of the first candidate responses, and determining, for the triggering event, whether the list of candidate responses includes a candidate response that has a confidence score that meets a triggering threshold. The method also includes waiting for a next triggering event without providing a candidate response when the list does not include a candidate response that has a confidence score that meets the triggering threshold.

Abstract: A knit sock is integrally formed of a body yarn in circumferential courses and axial wales, and includes a foot portion having a heel and a toe, and an ankle portion joined to the foot portion and defining an open leg end of the sock. A first targeted compression zone is formed with the foot portion between the heel and the toe, and includes spaced-apart axial compression ridges extending axially outward from the heel at a first end of the first compression zone to a second end of the first compression zone. A second targeted compression zone is formed with the ankle portion between the heel and the open leg end of the sock, and includes spaced-apart axial compression ridges extending axially outward from the heel at a first end of the second compression zone to a second end of the second compression zone.

Abstract: A method for automatically inserting hyperlinks is provided. In one example, the method includes determining a location for a hyperlink anchor in a document based on a type of structural element identified in the document. A presentation of the hyperlink anchor may be displayed according to a set of rules defining an appearance of the hyperlink anchor according to the type of structural element.

Abstract: The various embodiments described herein include methods and systems for encoding and decoding video. In one aspect, a method includes obtaining video data that includes a first block. The method further includes determining a plurality of transform coefficients associated with the first block. The method also includes predicting respective coefficient signs for the plurality of transform coefficients. The method also includes entropy coding the first block including jointly signaling two or more of the predicted coefficient signs.

Abstract: A method may include retrieving, from a geographic database, geographic information of a geographic area, segmenting the geographic area into one or more regions, executing a buildability algorithm using as input the geographic information to determine buildable portions of the one or more regions, executing a PV array placement algorithm using as input a total area of the buildable portions of the one or more regions to determine a baseline PV capacity for the one or more regions, executing the PV array placement algorithm using as input the buildable portions of the one or more regions to determine an actual PV capacity for the one or more regions, and generating a data structure to associate the one or more regions with a respective ratio of the baseline PV capacity to the actual PV capacity for each of the one or more regions.

Abstract: A method includes receiving training data that includes unspoken text utterances, un-transcribed non-synthetic speech utterances, and transcribed non-synthetic speech utterances. Each unspoken text utterance is not paired with any corresponding spoken utterance of non-synthetic speech. Each un-transcribed non-synthetic speech utterance is not paired with a corresponding transcription. Each transcribed non-synthetic speech utterance is paired with a corresponding transcription. The method also includes generating a corresponding synthetic speech representation for each unspoken textual utterance of the received training data using a text-to-speech model. The method also includes pre-training an audio encoder on the synthetic speech representations generated for the unspoken textual utterances, the un-transcribed non-synthetic speech utterances, and the transcribed non-synthetic speech utterances to teach the audio encoder to jointly learn shared speech and text representations.

Abstract: A system for load management in a shared address networking architecture includes a primary point-of-presence (POP) group of servers configured to serve content of a domain and that are each reachable at an address of a first IP address block and a secondary PoP group of servers configured to serve the content of the domain and that are each reachable at an address of a second IP address block. The system further includes a traffic management agent configured to reduce a total volume of incoming requests received by the primary PoP group for a period of time following a return of a first server in the primary PoP group of servers to an online state by selectively directing a first percentage of the incoming requests to the second IP address block instead of the first IP address block.

Abstract: An example apparatus includes a first 3D printer head configured to form a spiral structure around a hub and a second 3D printer head configured to form a boom extending between the second 3D printer head and the hub. The apparatus further includes one or more actuators coupled to the first 3D printer head and the second 3D printer head to control a distance between the first 3D printer head and the second 3D printer head.

Abstract: Generally discussed herein are devices, systems, and methods for digital signature generation security. A method can include generating, by a first device, a first random number, in generating a signature for a communication, masking, using the first random number, only a private key, a hash of the communication, or a combination thereof, and providing the signature with the communication to a second device.

Abstract: A block is provided that includes a first side having a first predetermined load-bearing area and a first non-load-bearing area, the first predetermined load-bearing area having a first predetermined size and a first predetermined location. A second side having a second predetermined load-bearing area and a second non-load-bearing area, the second predetermined load-bearing area having a second predetermined size and a second predetermined location. Wherein the first non-load-bearing area does not extend beyond the first predetermined load-bearing area.

Abstract: A computing system and method can be used to implement a version of federated learning (FL) that incorporates adaptivity (e.g., leverages an adaptive learning rate). In particular, the present disclosure provides a general optimization framework in which (1) clients perform multiple epochs of training using a client optimizer to minimize loss on their local data and (2) a server system updates its global model by applying a gradient-based server optimizer to the average of the clients' model updates. This framework can seamlessly incorporate adaptivity by using adaptive optimizers as client and/or server optimizers. Building upon this general framework, the present disclosure also provides example specific adaptive optimization techniques for FL which use per-coordinate methods as server optimizers. By focusing on adaptive server optimization, the use of adaptive learning rates is enabled without increase in client storage or communication costs and compatibility with cross-device FL can be ensured.

Abstract: A position sensor and a position sensor system, including a feeler, a force transfer element operably connected to the feeler, a force receiver operably connected to the force transfer element, and a force quantifier operably connected to the force receiver. A method for sensing position of a wet connector in a downhole environment, including running a position sensor on an upper wet connector to a target location, contacting a feeler block location on a lower wet connector, stroking the feeler, receiving a force at the force receiver, and communicating the received force to the force quantifier. A borehole system including a borehole in a subsurface formation, a position sensor disposed in the borehole.

Abstract: Methods and apparatus for prioritizing data traffic in small-cell wireless system. In one embodiment, the system utilizes “quasi-licensed” CBRS (Citizens Broadband Radio Service) wireless spectrum in conjunction with a wireless communication network (e.g., 3GPP 4G-LTE or 5G-NR-based). A “traffic profile” message is used to prioritize the data traffic. In one variant, the traffic profile message is generated by a Customer Premises Equipment (CPE) such as a CBRS FWA, and includes data related to CPE identification, default bearer identification, number of client devices served by the CPE (e.g., via a wireless router connected to the CPE), traffic type and/or priority level associated with each client device and/or data traffic type, etc. In another variant, the base station provides data services to a plurality of fixed wireless access (FWA) apparatus installed at user or subscriber premises of a network based on prioritization according to the present disclosure.

Abstract: Computer systems and methods train a deep neural network through machine learning. In response to detection of a training condition, computer system replaces a target node of the network with a split detector compound node, where, prior to replacement, the target node detected a pattern that activated the target node beyond a specified threshold. The split detector compound node comprises first and second nodes, such that: the first node is activated when significant evidence exists in favor of detection of the pattern in inputs to the first node; and the second node is activated when significant evidence exists against detection of the pattern in inputs to the second node, such that activations of the first and second nodes are computed independently. After replacing the target node with the split detector compound node, training of the network through machine learning is resumed.

Abstract: A medical clip that includes a pair of longitudinally extending legs that each have a tissue piercing distal end, and an interconnecting span section that connects proximal ends of the respective pair of longitudinally extending legs. The pair of longitudinally extending legs are disposed about a longitudinal axis; and further including a carrier element for supporting a radio-active source. The interconnecting span section is in the form of a platform having an elongated support surface along which the carrier element extends. The elongated support surface of the platform extends substantially orthogonal to the longitudinal axis defined between the pair of longitudinally extending legs.

Abstract: Disclosed herein are apparatus for a cooler device includes a manifold constructed at least partially of silicon oxide. The manifold includes an array of inlet channels and an array of outlet channels. Each inlet channel has a first depth. The array of outlet channels are interlaced with the array of inlet channels, Each outlet channel has a second depth that is larger than the first depth. Each of the array of outlet channels has a pair of sidewalls separating each outlet channel from adjacent inlet channels of the array of inlet channels.

Abstract: Apparatus and associated methods relate to a modular vehicle roof frame with at least two support members having at an upper end at least one coupling member through which the support member is axially assembled and releasably coupled to a plurality of top frame members such that each top frame member is releasably coupled to a plurality of support members. In an illustrative example, when the frame is assembled the support members may be substantially vertical and the top frame members may be substantially horizontal. When the frame is coupled to a vehicle the coupling members may, for example, be oriented substantially parallel to a longitudinal axis of the vehicle. The frame may be assembled, for example, by alternately assembling top frame members and support members. Various embodiments may advantageously allow a disassembled frame to be stored within a small sized container suitable for various ground shipping services.

Abstract: There is disclosed in one example a computing apparatus, including: a hardware platform including a processor and a memory; a network interface; a user-space application including instructions to interact with a web site via a uniform resource locator (URL); and a security agent including instructions to: intercept an interaction of the user-space application with the web site; determine that the intercepted interaction is to send sensitive information to the web site; suspend the interaction; and assign a reputation to the URL.

Abstract: A clip for a nut clip having a main body comprising an aft reaction boss defining at least a portion of a nut containment channel. The clip further includes a nut retention flange defining at least a further portion of the nut containment channel and connected to the main body by a nut retention flange attachment member spaced apart from the aft reaction boss. A forward reaction wall includes at least a portion of the nut retention flange attachment member facing the nut containment channel, wherein the aft reaction boss comprises an aft reaction wall. The aft reaction wall and the forward reaction wall are configured to be selectably contacted by a nut within the nut containment channel to occlude rotation of the nut.