Abstract: Systems and methods are provided for obtaining information from at least one computing system, the information including a set of records that respectively identify at least a network-based address of a computing device that accessed the computing system and an account hosted by the computing system that was accessed using the computing device; determining at least a first account and a second account were accessed from one or more computing devices that share a given network-based address based at least in part on the obtained information; and associating the first account and the second account with the network-based address.

Abstract: A method of creating a 3D model of a physical object includes adaptively and iteratively generating a number disparity maps from image data representing a plurality of images of the physical object iteratively captured by a plurality of cameras having electrically adjustable focal lengths by varying at least one of the focal lengths of the plurality of cameras and a distance of the physical object from the plurality of cameras during capture of the images until one of the disparity maps is determined to have a least a threshold level of disparity, and converting the one of the disparity maps into the 3D model.

Abstract: A system includes one or more source memory devices of a source computing environment that store a database comprising data files grouped in a plurality of file groups, wherein each of a plurality of data tables of the source computing environment includes data from one or more of the data files grouped into one or more of the file groups, one or more target memory devices of a target computing environment and at least one processor configured to calculate a number of read operations per megabyte (MB) for a data table, a number of write operations per MB, obtain an importance index of the table, weight the reads per MB, writes per MB and the importance index by respective user-defined weights, determine a criticality index of the table, calculate a value coefficient of the table, and assign a data file based on the value coefficient of the table.

Abstract: Filter media including fibers comprising a matrix polymer and impact modifier, and related methods, are generally described.

Abstract: A method for localizing a voice call is disclosed, comprising: receiving an originating leg setup message for an originating leg bearer from the first base station for a first user equipment (UE); creating a first call correlation identifier and storing the first call correlation identifier in association with the first UE; extracting a second call correlation identifier from a terminating leg setup message for a terminating leg bearer received from the core network; determining a real time protocol (RTP) localization status for the originating leg bearer and the terminating leg bearer based on matching the second call correlation identifier of the terminating leg against the stored first call correlation identifier of the originating leg; and sending transport layer assignment messages to the first base station to redirect RTP packets from the first UE to the second UE via the terminating leg bearer without the RTP packets transiting the core network.

Abstract: Directional pattern generation techniques are described for digital images as implemented by a directional pattern system. In an implementation, a user input is received to specify a direction with respect to the object. A directional pattern system then fills the object using a directional pattern based on the contours of the object as well as the user-specified direction. To do so, the directional pattern system generates a directional vector field that specifies directions with respect to corresponding locations within the field defined by a mesh. Uniform field embedding is employed to transfer the directional vector field to a grid by superimposing the grid onto the mesh of the directional vector field. The directional pattern system then generates the directional pattern within the object by filling the grid with one or more pattern cells.

Abstract: Systems and methods for providing notification delivery based on utilization of bloom filters are provided. A collaboration system obtains subscriber information for each user of a collaboration system, whereby the subscriber information corresponds to one or more features of content that are relevant to each user. The collaboration system hashes the subscriber information to generate a bloom filter for each user. The collaboration system receives an article to be published, whereby the article comprises a set of features. The set of features is hashed to obtain a hash set. The hashing of the set of features is performed using same hashing functions as that used to generate the bloom filter. The collaboration system compares the hash set to the bloom filter to identify a match, whereby the match indicates a feature of the article matches the subscriber information. The collaboration system generates a list of recipients based on the match.

Abstract: The invention relates to a headrest for a vehicle seat, comprising a first part; a second part that is able to rotate with respect to the first part; and a control rod for the relative movement between the first part and the second part; wherein the second part comprises an arcuate opening for receiving the rod, the opening comprising cavities; and wherein the rod is movable between an adjustment position in which the rod does not prevent the relative rotational movement between the first part and the second part; and a locked position in which the rod is trapped inside one of the cavities of the opening, thus preventing relative rotational movement between the first part and the second part.

Abstract: The present invention provides stable pharmaceutical compositions, comprising a pharmaceutically effective amount of testosterone or a pharmaceutically acceptable ester thereof, a pharmaceutically acceptable oil vehicle, and a stabilizing amount of benzyl alcohol, for example, about 1% to 3% weight/volume of benzyl alcohol. The present invention also provides a process for stabilizing testosterone-containing pharmaceutical compositions by ageing them at a temperature of about 20° C. to about 60° C. for at least 48 hours, e.g., prior to secondary packing and labeling. These compositions were stable over the shelf life of the product, without exhibiting crystal formation, even upon storing at temperatures of about 2° C. to about 8° C. Other aspects of the invention relate to methods for making such pharmaceutical compositions, as well as methods of using such pharmaceutical compositions for hormone replacement therapy, e.g.

Abstract: A machine implemented method includes alternately energizing multiple transmit coils in a first device, receiving indications of received signal strength at receive coils in a second device, selecting a first pair of coils including a first transmit coil and a first receive coil having the greatest received signal strength, and transferring energy from the first transmit coil to the first receive coil.

Abstract: Methods, systems, and computer programs are presented for suggesting related search queries. One method includes an operation for obtaining a supervised model by training a machine-learning (ML) program with training data that includes search queries entered by users of an online service. Further, the method includes operations for initializing a generator model with the supervised model, and for improving the generator model using reinforcement learning. The reinforcement learning is being based on a reward based on naturalness, relatedness, and a user having a positive session on the online service. Further, the result of the improvement of the generator model is a roll-out model, which is utilized to generate query suggestions for a user of the online service based on a search query provided by the user.

Abstract: This disclosure relates to a system that includes a centralized trim controller and a non-volatile memory that includes a trim sector configured for hosting trim data for one or more peripherals. The trim controller is configured to receive, for each of the one or more peripherals, trim values of the one or more peripherals from the trim sector of the nonvolatile memory, and provide the trim values to the one or more peripherals. Some trim values are updateable by receiving a password at the trim controller. If the password is valid, a timeout counter is initiated, during which time the trim value is updateable.

Abstract: Provided herein are methods for preparing ketone-containing organic molecules. The methods are based on novel iron/copper-mediated (“Fe/Cu-mediated”) coupling reactions. The Fe/Cu-mediated coupling reaction can be used in the preparation of complex molecules, such as halichondrins and analogs thereof. In particular, the Fe/Cu-mediated ketolization reactions described herein are useful in the preparation of intermediates en route to halichondrins.

Abstract: A memory circuit includes an array of memory cells arranged with first word lines connected to a first sub-array storing less significant bits of data and second word lines connected to a second sub-array storing more significant bits of data. A first word line signal is applied to a selected one of the first word lines to read less significant bits from the first sub-array, and a mathematical operation is performed on the read less significant bits to produce modified less significant bits that are written back to the first sub-array. If the read less significant bits are saturated, a second word line signal is applied to a selected one of the second word lines to read more significant bits from the second sub-array, and a mathematical operation is performed on the read more significant bits to produce modified more significant bits that are written back to the second sub-array.

Abstract: An DNN accelerator includes one or more heterogenous tile sets. A heterogenous tile set includes tiles of different sizes, e.g., PE arrays including different numbers of columns or rows. The DNN accelerator may identify a tile set from the tile sets for running a DNN model based on dimensions of output tensors convolutional layers in the DNN. Within the selected tile set, a tile may be selected for a convolutional layer in the DNN, e.g., based on dimensions of the output tensor of the convolutional layer and the size of the tile. After the tile is selected, the workload for running a convolutional operation of the layer may be partitioned and assigned to individual PEs in the tile by partitioning the output tensor into output tensor segments. The workload of computing an individual output tensor segment can be assigned to an individual PE in the tile.

Abstract: A computer-implemented method including receiving an input image at a first image stage and receiving a request to generate a plurality of variations of the input image at a second image stage. The method including generating, using an auto-regressive generative deep learning model, the plurality of variations of the input image at the second image stage and outputting the plurality of variations of the input image at the second image stage.

Abstract: A method, a device, and a non-transitory storage medium are described, which provide a natural language-based content system with corrective feedback and training service. The natural language-based content system with corrective feedback and training service may collect data based on interaction with search results from users. The natural language understanding model may generate feedback data based on the collected data, and use the feedback data to further train the natural language understanding model and update search and discovery logic for searching and discovering contents. The feedback data may categorize errors based on the interaction, and identify differences between search queries received during a search session with a user.

Abstract: In implementations of systems for generating stroked paths, a computing device implements a stroked path system to receive input data describing a vector object having a filled path. The stroked path system generates a medial axis for the filled path by performing a medial axis transform on a boundary of the filled path. A stroke width is estimated based on distances between the medial axis and the boundary of the filled path that are normal to the medial axis. The stroked path system generates a stroked path for display in a user interface that is visually similar to the filled path based on the medial axis and the stroke width.

Abstract: The present invention discloses a photocatalytic process and methodology for the preparation of acetic acid by a hydrocarboxylation reaction of methanol using carbon dioxide under visible light irradiation. Importantly, the reaction occurred under ambient temperature and pressure condition using a readily available household LED lamp in the presence of a transition metal based molecular photocatalyst, homogeneous as well as supported to semiconductor support and a CO2-philic solvent without adding any external electron and proton donors.

Abstract: Systems, methods, and computer-readable media are described for performing real-time vehicular incident risk prediction using real-time vehicle-to-everything (V2X) data. A vehicular incident risk prediction machine learning model is trained using historical V2X data such as historical incident data and historical vehicle operator driving pattern behavior data as well as third-party data such as environmental condition data and infrastructure condition data. The trained machine learning model is then used to predict the risk of an incident for a vehicle on a roadway segment based on real-time V2X data relating to the roadway segment and/or vehicle operators on the roadway segment. A notification of a high risk of incident can then be sent to a V2X communication device of the vehicle to inform an operator of the vehicle.