Abstract: A system includes one or more memory devices having instructions stored thereon that, when executed by one or more processors, cause the one or more processors to perform operations including receiving building information model (BIM) data associated with a building comprising one or more building assets, identifying one or more BIM objects within the BIM, identifying one or more object relationships between the one or more BIM object, applying a semantic description to each of the one or more BIM objects and the one or more object relationships, and generating a hierarchy structure for the building based on the one of more BIM objects and the one or more object relationships. The one or more BIM objects are associated with the one or more building assets and the hierarchy structure includes the semantic description of each of the one or more BIM objects and the one or more object relationships.

Abstract: Presented herein are systems and methods for the employment of machine learning models for image processing. A method may include a capture of a video feed including image data of a document at a client device. The client device can provide the video feed to another computing device. The method can include, by the client device or the other computing device object recognition for recognizing a type of document and capturing an image exceeding a quality threshold of the document amongst the frames within the video feed. The method may further include the execution of other image processing operations on the image data to improve the quality of the image or features extracted therefrom. The method may further include anti-fraud detection or scoring operations to determine an amount of risk associated with the image data.

Abstract: Maintaining the security of biometric data is an utmost priority. Biometric data is secured using one or more techniques. With one technique, biometric input such as images of a user's palm is used to generate first primary data (PD). The original biometric input is deleted from temporary secure storage while the first PD is securely stored. The first PD may then be processed later to determine a second PD. The first PD may then be deleted, and the second PD subsequently used. With another technique, biometric input or a PD may be processed by a first model to determine first secondary data (SD) that is representative of features of a particular user within a first embedding space. Later the PD may be processed by a second model to determine a second SD in a second embedding space. The first SD is deleted, and the second SD subsequently used.

Abstract: A user equipment (UE) configured to receive a first cumulative acknowledgement (ACK) from a protocol stack of the UE at a first time, store the first cumulative ACK in a queue, receive a second cumulative ACK from the protocol stack at a second time, wherein the second time occurs after the first time, store the second cumulative ACK in the queue, discard the first cumulative ACK from the queue based on the second cumulative ACK being received after the first cumulative ACK and encapsulate the second cumulative ACK to form a radio link control (RLC) packet data unit (PDU) that is to be transmitted over the air.

Abstract: Disclosed are techniques for using ranging signals to determine a position of a pedestrian user equipment (P-UE). In an aspect, a UE receives a plurality of ranging signals transmitted by one or more UEs, measures one or more properties of each of the plurality of ranging signals, and calculates an estimate of the position of the P-UE based on the one or more properties of each of the plurality of ranging signals. In an aspect, the P-UE transmits a plurality of ranging signals, receives a first message and a second message from first and second vehicle UEs (V-UEs), the first and second messages including first and second estimated positions of the P-UE and associated first and second confidences, and calculates an estimate of the position of the P-UE based on the first estimated position, the first confidence, the second estimated position, the second confidence, or a combination thereof.

Abstract: Disclosed is a system that includes a plurality of regional cloud exchange platforms coupled to a distributed alert triaging engine. A system can include a first regional cloud exchange platform and a second regional cloud exchange platform, each of which includes a regional cloud services monitoring engine and a regional cloud exchange monitoring engine, and an alert triaging engine that provides a triaged alert, or portion thereof, to an appropriate audience.

Abstract: A memory device includes bitcells connected to wordlines and bitlines, and driver circuitry that updates the bitcells. The driver circuitry includes first transistors, and a first inverter device. The first transistors drive a bitcell of a memory device. The first inverter device is coupled to the first transistors, and drives the first transistors with a first control signal. The first inverter device includes first inverter circuitry and second inverter circuitry. The first inverter circuitry receives a first signal, a first voltage, and a second voltage differing from the first voltage, and generates a first inverted signal based on the first signal, the first voltage and the second voltage. The second inverter circuitry receives the first inverted signal, the second voltage and a third voltage differing from the second voltage, and generates the first control signal based on the first inverted signal, the third voltage and the second voltage.

Abstract: A data storage device and method for host-assisted improved error recovery using a correlation factor are provided. In one embodiment, the data storage device receives, from a host, an indication that data associated with a first logical address is correlated with data associated with a second logical address; determines a correlation factor based on a degree of correlation between the data associated with the first logical address and the data associated with the second logical address; and in response to the correlation factor being above a threshold: stores the data associated with the first logical address and the data associated with the second logical address in different regions of the memory having different bit error rates; and uses the data associated with the first logical address to assist in correcting an error in the data associated with the second logical address. Other embodiments are provided.

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: Techniques described herein relate to intelligent migration between personal area networks (PANs) of a wireless mesh network. In one example, a method includes receiving, by a node connected to a first PAN, a packet from a second PAN. Based on the packet, the node determines that a second instability factor related to stability of the second PAN differs by at least a margin from a first instability factor related to stability of the first PAN. The node transmits, to a neighboring node in the first PAN, a migration notification message indicating that the node is migrating and receives, from the neighboring node in the first PAN, a request for the node to remain in the first PAN. The node applies one or more rules to reject the request for the node to remain in the first PAN and, based on rejecting the request, proceeds to complete the migration.

Abstract: A system configured to improve audio processing by performing dereverberation and noise reduction during a communication session. In some examples, the system may include a deep neural network (DNN) configured to perform speech enhancement, which is located after an Acoustic Echo Cancellation (AEC) component. For example, the DNN may process isolated audio data output by the AEC component to jointly mitigate additive noise and reverberation. In other examples, the system may include a DNN configured to perform acoustic interference cancellation, which may jointly mitigate additive noise, reverberation, and residual echo, removing the need to perform residual echo suppression processing. The DNN is configured to process complex-valued spectrograms corresponding to the isolated audio data and/or estimated echo data generated by the AEC component.

Abstract: Systems and methods for protecting and interacting with data in real time are described by the disclosed subject matter. A method includes monitoring a presentation of data by an application on a display and determining that a portion of the data is restricted. The method includes asking, in real time, the presentation of the portion of the data that is restricted.

Abstract: A monitoring system incorporates, and method and computer program product provide a monitoring system that captures image stream of regions of interest based on identified facial cues of a person. A controller of the monitoring system receives a first image stream that encompasses a face of a person and a second image stream that encompasses surrounding object(s) and surface(s) viewable by the person. The controller detects a facial expression of the face. In response to determining that the facial expression is a mood associated expression, the controller determines, from the first image stream, an eye gaze direction of the person. The controller determines a region of interest (ROI) aligned with the eye gaze direction. The controller captures the second image stream and identifies an object contained within the ROI. The controller communicates a notification including the expression and the object within the ROI to an output device.

Abstract: Digital representation techniques of intertwined vector objects are described. These techniques support a non-destructive representation of intertwined digital objects. Additionally, these techniques support editing of overlaps to change a visual ordering in an intuitive and efficient manner. Optimization operations are also implemented that remove redundancy, combine overlaps into a single representation, address visual artifacts at borders between the intertwined objected, and so forth.

Abstract: A method for performing a backup operation, the method comprising receiving a backup operation request for an asset, identifying a queue comprising the plurality of slices, wherein each slice references a separate portion of the asset, sending a first backup request to a proxy manager to instantiate a container for each of a plurality of backup sessions, wherein each backup session corresponds to a slice of the plurality of slices, receiving, from the proxy manager, a notification that one of the number of backup sessions is complete and a corresponding container has been torn down, making a second determination that there is an additional slice on a second queue associated with a second asset, and sending, based on the second determination, a backup request to the proxy manager to instantiate a new container for the additional slice associated with the second asset.

Abstract: A method provides non-mirrored preview of text-based demonstration objects in a locally captured video stream. The method includes identifying a demonstration object within the video stream. The demonstration object contains content that is best when pre-viewed in a non-mirrored orientation. The method includes spatially segmenting the video image into two or more segments, including a demonstration object preview (DOP) segment that encompasses the defined area with the demonstration object and at least one second segment encompassing a remaining portion of the video image. The method includes presenting a preview of the video image on the display device with the remaining portion of the video image mirrored within the preview and at least the DOP segment presented without mirroring in a correct spatial location relative to the remaining portion of the video image. A person presenting the demonstration object receives a non-mirrored presentation preview of the demonstration object.

Abstract: One or more engine instances are executed on each host to form an engine cluster. A plurality of control instances are executed on a first set of hosts to form a control cluster and comprise a control instance leader and one or more control instance followers. In response to a first host indicating a failure of a neighbor host, a pair-wise focused investigation is initiated to check peer-to-peer connections between the first host and the neighbor host. In response to one or more additional hosts indicating failures of neighbor hosts while the pair-wise focused investigation is being performed, a wide investigation is performed to check connections between the control cluster and the plurality of hosts. One or more hosts are added to an eviction list and an eviction protocol is performed to evict the one or more hosts from the engine cluster using the eviction list.

Abstract: Described herein is a latex coating composition comprising particles of polymer and articles therefrom, wherein at least 60% of interpolymerized monomeric units of the polymer are of formula I, wherein R3 is H or CH3, L is an alkylene group comprising 1-4 carbon atoms, R1 and R2 are independently selected from an alkyl group comprising 1 to 3 carbon atoms, n is an integer from 11 to 21, and A is a counter anion.

Abstract: A method for performing a backup operation, the method comprising receiving a backup operation request for an asset, partitioning a plurality of slices of the asset into a plurality of backup sessions, wherein each backup session comprises a separate portion of the plurality of slices, sending a first backup request to a proxy manager to initiate a backup session of the plurality of backup sessions, wherein initiating the backup session comprises: instantiating a container on the one compute node, wherein the backup session operates within the container, receiving a notification that one of the plurality of slices in the backup session has been processed, and sending, based on the notification, a second backup request to the proxy manager to add a new slice to the backup session, wherein the new slice is associated with a second asset.

Abstract: Methods and apparatus for duty cycle adjuster optimization training algorithms to minimize jitter associated with DDR5 DRAM transmitters. Basic and Advanced Duty Cycle Adjuster (DCA) training algorithms are implemented to reduce duty cycle error and hence reduce phase mismatch translated jitter in the transmitter DQS signals. In accordance with aspects of the Basic DCA training algorithm, duty cycles for QCLK, IBQCLK, and QBCLK are adjusted by a memory controller that utilizes a DCA function implemented on a memory module (e.g., DDR5 SDRAM DIMM) to obtain a first set of optimized DCA code settings. The first set of optimized DCA code settings are then used as initial settings for the Advance DCA training algorithm to further optimize the DCA code settings for QCLK, IBQCLK, and QBCLK. A similar technical may be employed to reduce duty cycle error and jitter for DQ signals.