Abstract: A method for providing elastic columnar cache includes receiving cache configuration information indicating a maximum size and an incremental size for a cache associated with a user. The cache is configured to store a portion of a table in a row-major format. The method includes caching, in a column-major format, a subset of the plurality of columns of the table in the cache and receiving a plurality of data requests requesting access to the table and associated with a corresponding access pattern requiring access to one or more of the columns. While executing one or more workloads, the method includes, for each column of the table, determining an access frequency indicating a number of times the corresponding column is accessed over a predetermined time period and dynamically adjusting the subset of columns based on the access patterns, the maximum size, and the incremental size.

Abstract: Shielded electrical transformers and power converters using those transformers are disclosed. In some implementations, a shielded electrical transformer includes a magnetic core, a primary winding, a first secondary winding, and a second secondary winding. The transformer includes a first shielding winding that has a same voltage potential direction as the primary winding and is connected in series with the primary winding to carry current that passes through the primary winding. The transformer also includes a second shielding winding that has a voltage potential direction opposite the primary winding and is connected from primary ground to a floating terminal. The first secondary winding, the second secondary winding, the first shielding winding, and the second shielding winding can each have an approximately equal number of turns.

Abstract: Techniques of delivering audio in a telepresence system include specifying a frequency threshold below which crosstalk cancellation (CC) is used and above which VBAP is used. In some implementations, such a frequency threshold is between 1000 Hz and 2000 Hz. Moreover, in some implementations, the improved techniques include modifying VBAP for more than three loudspeakers by forming an over-determined system to determine the amplitude weights for all loudspeakers at once.

Abstract: Implementations disclosed herein are directed to ephemeral learning of machine learning (“ML”) model(s) based on gradient(s) generated at a remote system (e.g., remote server(s)). Processor(s) of the remote system can receive stream(s) of audio data capturing spoken utterance(s) from a client device of a user. A fulfillment pipeline can process the stream(s) of audio data to cause certain fulfillment(s) of the spoken utterance(s) to be performed. Meanwhile, a training pipeline can process the stream(s) of audio data to generate gradient(s) using unsupervised learning techniques. Subsequent to the processing by the fulfillment pipeline and/or the training pipeline, the stream(s) of audio data are discarded by the remote system. Accordingly, the ML model(s) can be trained at the remote system without storing or logging of the stream(s) of audio data by non-transient memory thereof, thereby providing more efficient training mechanisms for training the ML model(s) and also increasing security of user data.

Abstract: Systems and methods for generating prompts for user data entry can include obtaining context data. The context data can be processed to determine whether an input entry interface is to be provided. In response to determining an input entry interface is to be provided, the context data or other data associated with a content display instance can be processed with a generative model to generate a prompt that can be provided to the user. User input data can then be obtained and stored to be provided to other users.

Abstract: Systems and methods that provide audio feedback in response to gesture validity can provide a more intuitive interface that can train users to correctly complete gestures. Moreover, systems and methods that provide line-specific audio feedback can provide more specific feedback that can allow a user to better understand what sensing line is being contacted. The systems and methods can further include basing the audio feedback based at least in part on obtained activity data, such that invalid and valid feedbacks can provide different sounds dependent on the determined activity state.

Abstract: The present disclosure provides a wearable device capable of evacuating fluids. The wearable device includes a component with an aperture, a bae port, and a bypass port. The base port extends outward from the aperture of the component to define a first opening on the wearable device. The bypass port extends outward from the base port to define a second opening on the wearable device. The present disclosure also provides methods for manufacturing a wearable device capable of evacuating fluids.

Abstract: A method, system and computer program product for workflow style governance includes creating a document and specifying a document style of the document—either a linear workflow style or a branched workflow style. Thereafter, in response to a subsequent request to load the document, the specified document style for the document may be retrieved and, on condition that the specified document style is a linear workflow style, a specified version of the document is retrieved into a document editor for editing as a new version of the document. But otherwise, on condition that the specified document style is a branched workflow style, different components of the document are retrieved as a composition of components, each of an independent version, and a selected one of the components of the composition of the components making up the document is then loaded into a component editor for editing as a new version.

Abstract: A method is performed at a server system. The server system receives from an electronic device media content information. The server matches the media content information with stored content information, and identifies a media content item based on the matching. It also identifies a genre of the media content item and one or more entities relevant to the identified media content item. The server generates one or more user-selectable elections based on the genre of the media content item. It sends to the electronic device the user-selectable elections and a plurality of affordances. Each of the affordances provides a user of the electronic device with one of the user-selectable elections corresponding to a respective entity that is relevant to the identified media content item. The server receives from the electronic device a selection of the one or more user-selectable elections, and performs an operation associated with the selection.

Abstract: A computing system can be configured to input model input that includes context data into a machine-learned model and receive model output that describes one or more semantic entities referenced by the context data. The computing system can be configured to provide data descriptive of the semantic entity or entities to the computer application(s) and receive application output(s) respectively from the computing application(s) in response to providing the data descriptive of semantic entity or entities to the computer application(s). The application output(s) received from each computer application can describe available action(s) of the corresponding computer application with respect to the semantic entity or entities. The computing system can be configured to provide at least one indicator to a user that describes the available action(s) of the corresponding computer applications with respect to the semantic entity or entities.

Abstract: The systems, methods, and techniques described in this disclosure allow different wireless systems that operate in accordance with different Radio Access Technologies (RATs) to coexist within a same frequency domain with minimal (if any) inter-RAT interference. Specifically, the described techniques allocate a respective, mutually-exclusive portion of a plurality of Space-Time-Frequency (STF) resources for use in communicating in accordance with each different RAT. For example, mutually-exclusive portions of spatial domain resources, time domain resources, and/or frequency domain resources may be respectively allocated for exclusive use by different RATs. A centralized, third-party controller (120) may perform the allocations, or the allocations may be cooperatively arrived at between systems supporting different RATs, e.g., in a peer-to-peer manner. STF resource allocations may be static and/or dynamic over time, and STF resources may be uniquely identified by respective resource identifiers.

Abstract: A method of detecting and responding to a visitor to a smart home environment via an electronic greeting system of the smart home environment, including determining that a visitor is approaching an entryway of the smart home environment; initiating a facial recognition operation while the visitor is approaching the entryway; initiating an observation window in response to the determination that a visitor is approaching the entryway; obtaining context information from one or more sensors of the smart home environment during the observation window; and at the end of the time window, initiating a response to the detected approach of the visitor based on the context information and/or an outcome of the facial recognition operation.

Abstract: A haptic system can include a user device including one or more haptic actuators and one or more sensors. The haptic system can be configured to perform operations including controlling the haptic actuator(s) to produce a haptic output based on a haptic output profile associated with a desired action of a user of the haptic system; detecting, using the one or more sensors, one or more user response characteristics in response to the haptic output; determining, based at least in part on the one or more user response characteristics, data describing a user action of the user performed after the haptic actuator(s) produces the haptic output; comparing the data describing the user action with data indicative of the desired action to generate an effectiveness metric; and determining a subsequent haptic output profile for a subsequent haptic output based at least in part on the effectiveness metric.

Abstract: Methods, systems, devices, and tangible non-transitory computer readable media for surface detection and geolocation are provided. The disclosed technology can receive location request data associated with a request for a geographic location of a user device in an environment. Based on the location request data, images of segments of a surface of the environment can be accessed. The geographic location of the user device can be determined based on the segment signatures that match stored segment signatures respectively associated with stored geographic locations. The segment signatures can be based on the images of the segments of the surface of the environment. Map data including information associated with a geographic area can be accessed. Furthermore, indications can be generated and the indications can include information associated with the geographic location of the user device within the geographic area.

Abstract: A multiple graphics processing unit (GPU) based parallel graphics system comprising multiple graphics processing pipelines with multiple GPUs supporting a parallel graphics rendering process having an object division mode of operation. Each GPU comprises video memory, a geometry processing subsystem and a pixel processing subsystem. According to the principles of the present invention, pixel (color and z depth) data buffered in the video memory of each GPU is communicated to the video memory of a primary GPU, and the video memory and the pixel processing subsystem in the primary GPU are used to carry out the image recomposition process, without the need for dedicated or specialized apparatus.

Abstract: A neural network system that includes: multiple subnetworks that includes: a first subnetwork including multiple first modules, each first module including: a pass-through convolutional layer configured to process the subnetwork input for the first subnetwork to generate a pass-through output; an average pooling stack of neural network layers that collectively processes the subnetwork input for the first subnetwork to generate an average pooling output; a first stack of convolutional neural network layers configured to collectively process the subnetwork input for the first subnetwork to generate a first stack output; a second stack of convolutional neural network layers that are configured to collectively process the subnetwork input for the first subnetwork to generate a second stack output; and a concatenation layer configured to concatenate the pass-through output, the average pooling output, the first stack output, and the second stack output to generate a first module output for the first module.

Abstract: A method of configuring an integrated circuit including multiple hardware tiles, includes: establishing a data forwarding path through the multiple hardware tiles by configuring each hardware tile, except for a last hardware tile, of the multiple hardware tiles to be in a data forwarding state, in which configuring each hardware tile, except for the last hardware tile, to be in a forwarding state includes installing a respective forwarding state counter specifying a corresponding predefined length of time that the hardware tile is in the data forwarding state; supplying, along the data forwarding path, each hardware tile of the plurality of hardware tiles with a respective program data packet comprising program data for the hardware tile; and installing, for each hardware tile of the multiple hardware tiles, the respective program data.