Abstract: Exemplary embodiments relate to improvements in the design of a messaging inbox. The inbox may display different units or “modules” for providing a user with quick access to different inbox functionalities. After a number of recent or unread messages are shown in the inbox's initial interface, the messages end and are replaced with modules. A threshold may be defined for the number of recent/unread messages to display before transitioning to modules. The threshold may be determined dynamically based on a minimum and/or maximum number of messages to display. The determination may be based on the current time, whether there is an active conversation in a thread, whether there are unread messages in a thread, etc. The determination may vary from user to user based, for example, on whether the user is a relatively active user, and/or how the user has used the messaging application in the past.

Abstract: Techniques for eye-tracking in a near-eye display system are disclosed. One example of a near-eye display system includes a substrate transparent to visible light and configured to be placed in front of a user's eye, one or more light sources configured to emit illumination light invisible to the user's eye, one or more input couplers configured to couple the illumination light into the substrate, and one or more grating couplers each configured to couple a portion of the illumination light out of the substrate and towards the user's eye at a different direction. The illumination light coupled into the substrate propagates within the substrate through total internal reflection. The one or more grating couplers include at least one of a chirped surface-relief grating coupler or a volume Bragg grating coupler.

Abstract: Systems and methods for fabricating parabolic LEDs are provided. The parabolic shape of the LEDs is precisely-controlled and highly-uniform across a substrate. By precisely controlling the shape, and providing a high-uniformity across the substrate, the luminance and process yield of the LEDs is enhanced. The precise-control and high-uniformity of the shape is enabled via a precisely-shaped and highly-uniform mask formed on the substrate. The ability to precisely-control both the shape and uniformity the mask is achieved by forming the mask utilizing three-dimensional (3D) patterning and/or machining methods. The mask includes a precisely-shaped boss with the same shape as the LED, and a cylindrical protrusion extending beyond the boss. The combination of the boss and cylindrical protrusion allows for the mask to be over-etched, without significantly effecting the shape of the LED. Thus, any non-uniformities etching process do not decrease the luminance, nor uniformity, of the LEDs.

Abstract: The present disclosure relates to the manufacture of battery packs/assemblies and more specifically, the manufacture of battery packs/assemblies for use in aircraft. A lightweight battery assembly with cell compression and/or pressure management system is disclosed herein. The cell compression and/or pressure management system can employ a sealed container housing a volatile vapor. The sealed container can be positioned within the battery enclosure and configured to impart a force to compress the cell-stack.

Abstract: A first waveguide emits first light and a second waveguide emits second light. A filter is configured to reflect the second light and pass the first light to an entrance facet of the second waveguide.

Abstract: In one embodiment, a method includes determining television content that a particular user is currently watching on a television. The method further includes determining, using one or more of a plurality of sensors, an attention level for the particular user. The attention level indicates an amount of attention paid by the particular user to the television content on the television. The method further includes determining, using one or more of the plurality of sensors, demographic information for the particular user. The method further includes storing the determined demographic information for the particular user and the particular user's attention level for the television content in the storage media.

Abstract: In one embodiment, a method includes preparing an initialized user-embedding matrix, where the user-embedding matrix includes multiple columns corresponding to a user, preparing an initialized concept-embedding matrix, constructing an interaction table representing interactions between entities, where an element (i,j) of the table indicates one or more directional interactions from an entity represented by row i to an entity represented by column j, running a training algorithm for the embeddings in the user-embedding matrix and in the concept-embedding matrix, where the training algorithm trains the embeddings such that embeddings representing entities with interactions are located nearby in the embedding space while embeddings representing entities without interactions are located far apart, and providing the embeddings in the user-embedding matrix and in the concept-embedding matrix to a downstream service.

Abstract: In one embodiment, a method includes sending a request from a first user for communication-channel information associated with a second user to a computing device, and receiving one or more user-selectable elements corresponding to each communication channel of a plurality of communication channels from the computing device, wherein the plurality of communication channels are determined based on an affiliation coefficient between the first user and the second user, wherein each communication channel is associated with a score calculated based on a social proximity between the first user and the second user, and wherein the one or more user-selectable elements corresponding to each communication channel of the plurality of communication channels are presented in an order ranked based on the respective score associated with the communication channel.

Abstract: An improved lecture support system integrates multi-media presentation materials with spoken content so that the listener can follow with both the speech and the supporting materials that accompany the presentation to provide additional understanding. Computer-based systems and methods are disclosed for translation of a spoken presentation (e.g., a lecture, a video) along with the accompanying presentation materials. The content of the presentation materials can be used to improve presentation translation, as it extracts supportive material from the presentation materials as they relate to the speech.

Abstract: In one embodiment, one or more main processors of a device detect an event corresponding to scrolling a structured document presented in a scrollable graphical user interface of an application executed by the one or more main processors. The main processors, in response to the event, determine a first set of elements and a second set of elements of the structured document; render the first set of elements of the structured document for presentation in the graphical user interface; and cause one or more graphics processors of the device to render the second set of elements for presentation in the graphical user interface.

Abstract: For a visit of a user to a web page where the user's identity on an online system is not presently known to the online system, the online system uses a machine learning model to make a prediction of the user's identity. The online system obtains visit data about the visit of the user to the web page. The online system identifies candidate user IDs that may represent the user, based on the visit data and data known about previous visits of the candidate user IDs. The online system derives visit features for each candidate user ID based on a relationship between the current visit data and previous visit data for the candidate user ID. The online system provides the visit features for each candidate user ID to a prediction model that determines whether, or how likely, the candidate user ID accurately identifies the visiting user, and based on the determinations selects one of the candidate user IDs as the most likely user ID for the visiting user.

Abstract: A display includes a display pixel array and a light bending assembly. The display pixel array is configured to generate display light. The light bending assembly is disposed over the display pixel array to receive the display light and generate compensated display light. The light bending assembly bends a given ray of the display light based on an incidence position that the given ray of the display light becomes incident upon the light bending assembly.

Abstract: Systems, methods, and non-transitory computer-readable media can determine one or more source video clips. A plurality of video segments are selected from the one or more source video clips based on video segment selection criteria. A compiled video is generated comprising the plurality of video segments.

Abstract: Various embodiments of a modular electromechanical device are described herein. The modular electromechanical device includes a chassis and a plurality of functional modules that can be connected to the chassis. Each module is associated with a different functionality. The functionality of the modular electromechanical device is defined based on various attributes including the functionality of the different functional modules that are connected to the electromechanical device, the sequence in which the different functional modules are connected to the electromechanical device, the specific attachment structures used to attach the functional modules to the electromechanical device, or a pattern of traces formed within the chassis.

Abstract: Exemplary embodiments relate to improvements to neural networks for translation and other sequence-to-sequence tasks. A convolutional neural network may include multiple blocks, each having a convolution layer and gated linear units; gating may determine what information passes through to the next block level. Residual connections, which add the input of a block back to its output, may be applied around each block. Further, an attention may be applied to determine which word is most relevant to translate next. By applying repeated passes of the attention to multiple layers of the decoder, the decoder is able to work on the entire structure of a sentence at once (with no temporal dependency). In addition to better accuracy, this configuration is better at capturing long-range dependencies, better models the hierarchical syntax structure of a sentence, and is highly parallelizable and thus faster to run on hardware.

Abstract: A digital audio gateway device for use in a wireless network of digital audio playback devices. The gateway device is wirelessly linked to one or more digital audio playback devices to provide a gateway to the Internet for the digital audio playback devices. In addition to functioning as a gateway, the device provides additional functionality and may act as a cache of digital audio data for the various digital audio players connected in the wireless network and may also act to automatically update digital audio content on the audio players, synchronize digital audio content and playlists between the digital audio players and continue automatically or upon user request a particular playlist as the user moves from one digital audio player to another.

Abstract: A chassis is configured to receive a movable component inserted in the chassis. A lever arm is housed at least in part within the chassis and configured to rotate about a pivot point from a force applied on the lever arm by a contact portion of the movable component inserted in the chassis. At least a portion of an electrical connector is coupled to the lever arm and configured to rotate and mate with a mating portion located on a lateral side of the moveable component.

Abstract: A fluidic-device valve having a valve guide with a guide wall surrounding a central axis. The guide wall may include a first guide section, a second guide section, and an interior surface that defines a valve cavity extending longitudinally along the central axis. The valve may also include a valve member disposed within the valve cavity, and the valve member may extend longitudinally along the central axis between a first end and a second end of the valve member. The valve member may be movable within the valve cavity in a first axial direction along the central axis to position a fluid conduit defined by the valve member at a first location by increasing a pressure applied to the second end of the valve member by a pressure source coupled to the second guide section. Various other apparatuses, systems, and methods are also disclosed.

Abstract: An expansion card may include a printed circuit board and a hardware accelerator that is disposed on the printed circuit board. The hardware accelerator may include application-specific hardware circuitry designed to perform a computing task. The hardware accelerator may also offload a portion of the computing task from a central processing unit of a computing device by executing, via the application-specific hardware circuitry, the portion of the computing task. The expansion card may further include an edge connector, disposed on the printed circuit board, that is dimensioned to be inserted into an expansion socket of the computing device. The edge connector may couple the hardware accelerator to the central processing unit via a computing bus connected to the expansion socket. The edge socket may also include a pinout that is more compact than a pinout specification defined for the computing bus. Various other apparatuses, systems, and methods are also disclosed.

Abstract: In one embodiment, a system includes a first projector configured to project a first projected pattern having one or more first projected lighting characteristics, a second projector configured to project a second projected pattern having one or more second projected lighting characteristics, a camera configured to capture an image comprising first and second detected patterns corresponding to reflections of the first and second projected patterns, respectively, and one or more processors configured to: identify a detected point in the image that corresponds to a projected point in at least one of the first and second projected patterns by comparing detected lighting characteristics of the first and second detected patterns with the first and second projected lighting characteristics, and compute a depth associated with the detected point based on the projected point, the detected point, and a relative position between the camera and at least one of the projectors.