Abstract: Systems and methods for geolocation-based network slice allocations for dynamic edge computing resource management are provided. In some embodiments, a network function of an operator core network may orchestrate one or more service augmentation platforms on network edge servers based on identifying event(s) where UE use the telecommunications network to access cloud-based service from an edge server within a threshold proximity of the event(s). Orchestration of the service augmentation platform(s) may be based in part on determining a network service classification for the event(s) based on event data collected about the event(s). A network slice for accessing the service augmentation platforms may be granted to UE by the network function based in part on the UE geographic proximity to the location associated with the scheduled events. The network function may de-orchestrate the service augmentation platform based on the end of the event as indicated by the event data.

Abstract: A computer-implemented system and method relate to gesture recognition. A machine learning system is trained using a training dataset of sensor data that include a set of gestures. The training dataset includes at least a first subset that displays a first gesture. Loss data is generated based on a first loss function that includes a first cross entropy loss and a second cross entropy loss. Parameters of the machine learning system are updated based on the loss data. The machine learning system is outputted and configured for gesture recognition of the set of gestures. The machine learning system includes (i) a first subnetwork to generate feature data based on the sensor data, (ii) a second subnetwork to extract a selected patch of the feature data, and (iii) a third subnetwork to generate gesture data based on a classification of the corresponding feature data of the selected patch. The first cross entropy loss is based on a first performance of the second subnetwork in relation to the training dataset.

Abstract: Machine learning techniques are leveraged to provide personalized assistance on a computing device. In some configurations a timeline of a user's interactions with the computing device is generated. For example, screenshots and audio streams may be saved as entries in the timeline. Context—the state of the computing device when the entry is created, such as which documents and websites are open—is also stored. Entries in the timeline are processed by a model to generate embedding vectors. The timeline may be searched by finding the embedding vector that is closest to an embedding vector derived from a search query. The user may select a query result, causing the associated context to be restored. For example, if the query is “show me all documents related to my upcoming trip to Japan”, the query result may open documents and websites that were open when booking a flight to Japan.

Abstract: A method includes determining that a user device is located within a specified geographical region, and determining that the user device has an active subscription with a terrestrial network operator operating a terrestrial network within the geographical region, the terrestrial network associated with a primary profile stored on a subscriber identity module of the user device. The method also includes, in response to determining that the user device is located within the geographical region and that the user device has an active subscription with the terrestrial network operator, receiving, from the terrestrial network operator, a secondary profile attached to the active subscription and associated with a non-terrestrial communication network. The method further includes storing the secondary profile on an embedded subscriber identity module on the user device, and communicating via the non-terrestrial communication network using the secondary profile.

Abstract: A computer-implemented method comprises: receiving first telemetry data generated by sensors of respective first vehicles in a fleet; clustering the first telemetry data into groups, each of the groups representing a profile of one or more first drivers of the first vehicles in the fleet; receiving second telemetry data generated by sensors of a second vehicle controlled by a second driver; associating the second driver with a first group of the groups by classifying the received second telemetry data; providing a subset of the first telemetry data corresponding to the first cluster as a baseline dataset for training of machine learning algorithms; generating baseline tuning parameter values using the trained machine learning algorithms; and providing the baseline tuning parameter values to a driver assistance system of a third vehicle controlled by the second driver.

Abstract: Systems and methods for dynamic utilization-based network slice allocation management for user equipment applications are provided. In some embodiments, a slice estimation engine may be implemented to evaluate the network traffic and other application activity data associated with an application running on the UE to determine an operating mode of the application. The slice estimation engine may trigger the UE to request an adjustment to its network slice allocation configurations based on the evaluation. To determine whether or not an application should be reconfigured for a new network slice, the slice estimation engine may evaluate processes that are running on the UE. The slice estimation engine may comprise one or more slice assessment algorithms that determine which slice from a set of available network slices would optimally serve the application based on the network traffic characteristics associated with the application's current mode of operation.

Abstract: The disclosed technology relates to methods, speech processing systems, and non-transitory computer readable media for style extraction in speech synthesis. In some examples, one or more content elements and one or more non-content elements are extracted from input audio data obtained via an audio interface and corresponding to input speech. The one or more non-content elements comprise style elements comprising at least an input pitch. A trained autoencoder is applied to encode the input pitch in a latent representation comprising a low-dimensional vector and combine the one or more content elements and the one or more non-content elements based on the low-dimensional vector to generate a new representation of the input speech. Output audio data is then generated and provided based on the new representation of the input speech. The output audio data comprises a pitch-consistent reconstruction of the input speech.

Abstract: The disclosed technology relates to methods, speech processing systems, and non-transitory computer readable media for style extraction in speech synthesis. In some examples, one or more content elements and one or more non-content elements are extracted from input audio data obtained via an audio interface and corresponding to input speech. The one or more non-content elements comprise style elements comprising at least an input pitch. A trained autoencoder is applied to encode the input pitch in a latent representation comprising a low-dimensional vector and combine the one or more content elements and the one or more non-content elements based on the low-dimensional vector to generate a new representation of the input speech. Output audio data is then generated and provided based on the new representation of the input speech. The output audio data comprises a pitch-consistent reconstruction of the input speech.

Abstract: Video and audio from a computer simulation are processed by a machine learning engine to identify candidate segments of the simulation for use in a video summary of the simulation. Text input is then used to reinforce whether a candidate segment should be included in the video summary.

Abstract: A fluid delivery system having a fluid delivery device configured to deliver a fluid in a spray form. The fluid may be a first fluid and a second fluid mixed together in a manner causing the first fluid to be delivered in a spray form. The first fluid may be a viscous fluid, such as a hemostatic agent, and the second fluid may be a pressurized fluid. The second fluid may be directed at the first fluid to cause the first fluid to be delivered in a spray form. The fluid delivery device may be adjustable to adjust the spray form. The lumens through the fluid delivery device and through the fluid delivery system are formed by channel walls and are uninterrupted by additional/separate components (e.g., valves, etc.) therein affecting fluid flow through the lumens.

Abstract: The system scans a RAN associated with a wireless telecommunication network to obtain an indication of a hardware status and a software status associated with the RAN. The system determines whether there is an update to the hardware or software status associated with the RAN. Upon determining there is the update, the system obtains from a first database an indication of a remote radio head configured to test the update. Based on the indication of the remote radio head, the system obtains from a second database multiple UEs configured to test the update. The system tests the update by causing the remote radio head to send a communication to a UE among the multiple UEs. The system obtains logs associated with the communication and, based on the logs, determines whether the update passes the tests.

Abstract: An end user premises device is provided that includes a memory, one or more transceivers, and one or more processors. The one or more transceivers are configured to communicate with one or more stations in a network and a client device. The one or more processors are configured to receive a first user request for data from the client device using the one or more transceivers, determine a first point in time for retrieving the data based on an amount of charge in batteries of the one or more stations in the network, retrieve, at the first point in time, the data from a remote server via the network using the one or more transceivers, store the data in the memory, and in response to a second user request, transmit the data to the client device using the one or more transceivers.

Abstract: A maritime network provides network coverage for nautical or aerospace vehicles traveling over the sea. Generating the network configuration for the maritime network includes receiving client information for client devices in range of a given node of the maritime network for a period of time that the client devices are traveling asea, as well as location information for the period of time from a plurality of nodes in the network including the given node. Based on the client information and the location information, a network configuration is determined to include a plurality of links to be formed for routing paths through the maritime network. The routing paths are configured to transmit data related to the client devices, and the plurality of links includes a link between the given node and another node in the network that is within a maximum distance from the given node.

Abstract: Apparatus and methods for improving optical signal collection in an integrated device are described. A microdisk can be formed in an integrated device and increase collection and/or concentration of radiation incident on the microdisk and re-radiated by the microdisk. An example integrated device that can include a microdisk may be used for analyte detection and/or analysis. Such an integrated device may include a plurality of pixels, each having a reaction chamber for receiving a sample to be analyzed, an optical microdisk, and an optical sensor configured to detect optical emission from the reaction chamber. The microdisk can comprise a dielectric material having a first index of refraction that is embedded in one or more surrounding materials having one or more different refractive index values.

Abstract: A computing device for collecting and preserving data acquired from various devices in a data model with the respective context of acquired data. The computing device may provide a user interface to receive the context or information model associated with a dataset. By providing data with its context, different software platforms may synthesize or analyze the retrieved data more efficiently. Moreover, the computing device may include transaction conditions to define a workflow for transferring the datasets using data model associated with one or more datasets for transmission of data to a destination. The transaction conditions may detail a custom workflow for data communication through an industrial automation system using the data model.

Abstract: An electrical terminal includes a main body that extends along a plane. The electrical terminal includes at least one shoulder extending away from the main body. The electrical terminal includes a flag snap positioned downward from the at least one shoulder. The electrical terminal can be installed within an electronic device. The electronic device includes a housing and a circuit board positioned within the housing.

Abstract: Described herein are component(s) and device(s) of a telecommunications network. One or more of the components is configured to detect impact at a cell site on designated spectrum by a neighboring country user equipment (UE) associated with a different telecommunications network. A base station of the component(s) and device(s) blanks physical resource blocks (PRBs) associated with the designated spectrum in response to the detecting.

Abstract: Inline Multimodal Data Loss Protection (DLP) includes training one or more machine learning models for classifying input data into categories of a plurality of categories; performing one or more modifications to the one or more machine learning models, wherein the one or more modifications reduce latency associated with the one or more machine learning models; receiving an input comprising data in any of a plurality of formats; processing the input to classify the input into a category of a plurality of categories; and providing an indication of the category of the plurality of categories. Advantageously, by performing the various modifications to the one or more models, the systems can accurately classify data inline with minimal latency.

Abstract: A processor-implemented method for a priority-based cache eviction policy includes receiving a read request from a first processing core for first data in a level three (L3) cache memory shared with a second processing core. The first processing core has a first operating frequency that is less than a second operating frequency of the second processing core. Responsive to the L3 cache being full, the policy further includes determining a second data or a third data stored in the L3 cache to evict based on a priority. The priority is based on an association of the second data or the third data to the first processing core or the second processing core.

Abstract: Some embodiments provide order packaging control systems, comprising: a product packaging (PP) control circuit communicatively couple with an automated custom package fabrication (CPF) system configured to fabricate on demand custom packages, an automated product retrieval (PR) system configured to autonomously retrieve one or more products, pick stations, and a set of sensor systems. The PP control circuit can be configured to receive orders; prioritize the orders and define an order fulfillment sequence, track states of the PR system and the CPF system; determine product retrieval timing, and correspondingly determine fabrication timing of when the CPF system is to fabricate corresponding packaging; and control the PR system and the CPF system in controlling synchronous operation of the PR system and the CPF system to synchronize an availability of products of an order at a pick station and an availability of a custom package at the pick station.