Abstract: An ear-plug assembly presents audio content to an ear canal of a user. The audio content may be based in part on sound in a local area surrounding the user. The ear-plug assembly detects, via one or more acoustic sensors, sound in the area around the user. The sound waves travel through an aperture in a body of the ear-plug assembly and are propagated to a waveguide to the one or more acoustic sensors. The ear-plug assembly processes the detected sound data in a controller, which instructs a speaker assembly to present audio content based in part on the detected sound data. The detected sounds may be amplified, attenuated, filtered, and/or augmented when presented by the speaker assembly.

Abstract: Various of the present embodiments convert structured data to a channel-based representation facilitating pattern recognition by a compression utility. The channels may include a key channel, a number channel, a type channel, a length channel, etc. These channels reflect the same information as the structured data, but reorganized based upon the structured data's hierarchic structure. The encoder and decoder may include rule sets specifying the relations between the channels to deconstruct/construct the JSON file. Memoization, length encoding, and type encoding extensions may also be applied as part of the encoding/decoding process to further improve compression.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods for dynamically transitioning a digital video file between encoding states of video data encodings. For example, upon receiving an uploaded digital video file, the disclosed systems can generate an initial set of video data encodings for the digital video file in an initial state. The disclosed systems further determine a virality score associated with the digital video file. Based on the virality score, the disclosed systems determine the virality score satisfies a threshold for transitioning the digital video file from the initial state to the upgraded state and generate premium encodings for the digital video file. Similarly, based on an updated virality score or a time-based trigger, the disclosed systems can transition the digital video file from the upgraded state (or the initial state) to the maintenance state by purging one or more video data encodings.

Abstract: In one embodiment, a method includes a system building a first machine-learning module and one or more secondary machine-learning modules for operating with an application. The first and second modules may be configured to utilize, in operation, particular types of processing hardware, respectively. The system may receive from a client device a request to download the application, and in response send to the device the application with the first module. The system may then receive another request to download a selected one of the secondary modules, which may be selected based on a determination by the application running on the device that the associated type of processing hardware is available. In response, the system may send the selected module to the device. The application may be configured to selectively use the first module or the selected module to perform an operation.

Abstract: The disclosed system may include (1) a drive subsystem that translates along a powerline conductor, (2) a rotation subsystem that rotates a segment of fiber optic cable about the powerline conductor while the drive subsystem translates along the powerline conductor such that the segment of fiber optic cable is wrapped helically about the powerline conductor, and (3) an extension subsystem that (a) mechanically couples the rotation subsystem to the drive subsystem, and (b) selectively extends the rotation subsystem away from the drive subsystem and the powerline conductor to avoid obstacles along the powerline conductor. Various other systems and methods are also disclosed.

Abstract: In one embodiment, a method includes a computing system accessing a content item associated with a content provider, the content item having a first set of attributes and a second set of attributes. The system may generate, using a first machine-learning model, a first ranking score of the content item for a user based on the first set of attributes. The system may generate cluster representations of the second set of attributes of the content item. The system may generate, using a second machine-learning model, a second ranking score of the content item for the user based on the cluster representations. The system may generate, using a third machine-learning model, a third ranking score of the content item for the user based on the first ranking score and the second ranking score. The system may select the content item for presentation to the user based on the third ranking score.

Abstract: In one embodiment, a method includes sending, to an interaction device including a projector and a camera, a media content item and instructions causing the projector to project the media content item on a projectable surface and receiving, from the interaction device, one or more media objects captured by the camera, where one or more of the media objects include images of a user in proximity to the projectable surface. The method includes determining one or more movements of the user based on the one or more of the media objects and updating the media content item based on the determined movements. The method also includes sending, to the interaction device, the updated media content item and instructions causing the projector to project the updated media content item on the projectable surface.

Abstract: Systems and techniques are described that enable users to interact and share content through a social network application and/or service with other users. A social networking system may determine that a first application and a second application are installed on a device and are both useable to share content between a first account and one or more contact accounts. The social networking system may receive content associated with the first application and the second application and may generate a notification associated with the content. The social networking system may determine which of the first application or the second application to use to present the notification, and may output the notification to the determined first application and/or the second application.

Abstract: Techniques for intelligent messaging for message syncing are described. An apparatus may comprise a recipient inbound messaging component, a recipient queue management component, a recipient update customization component, and a recipient outbound messaging component. The recipient inbound messaging component may be operative to receive an incoming update at a recipient update queue, the recipient update queue associated with a recipient of the incoming update. The recipient queue management component may be operative to add the incoming update to the recipient update queue and determine a recipient messaging endpoint to receive the incoming update. The recipient update customization component may be operative to retrieve one or more recipient messaging endpoint parameters associated with the recipient messaging endpoint and generate a customized incoming update from the incoming update according to the one or more recipient messaging endpoint parameters.

Abstract: Techniques for enhancing group communication on a device are described. A method may include presenting a message in a message portion of a user interface (UI) for a group communication application executing on a first device, where the group communication application communicates messages among multiple devices. The method may further include monitoring an input component for the first device and detecting a first control directive from the input component. The first control directive may select a message displayed in the message portion of the user interface. The method may further include performing a pin operation on the selected message; and presenting the pinned message in a banner portion of the UL Other embodiments are described and claimed.

Abstract: Techniques are described for content-adaptive downsampling of digital images and videos for computer vision operations, such as semantic segmentation. A computer vision system comprises a memory, one or more processors operably coupled to the memory and a downsampling module configured for execution by the one or more processors to perform, based on a non-uniform sampling model trained to predict content-aware sampling parameters, downsampling input image data to generate downsampled image data. A segmentation module is configured for execution by the one or more processors to perform segmentation on the downsampled image to produce a segmentation result, such as a feature map that assigns pixels of the downsampled image data to object classes. An upsampling module is configured for execution by the one or more processors to perform upsampling according to the segmentation result to produce upsampled image data.

Abstract: Systems, methods, and non-transitory computer readable media can determine a representation of an advertisement based on a first machine learning model. The representation can be provided to a second machine learning model. One or more qualitative ratings associated with the advertisement can be determined based on the second machine learning model.

Abstract: Techniques are described that determine a user to participate in software testing of a social network application and/or service and to enable the user to propagate an invitation to contacts of the user to participate in the software testing of the social network application. In some examples, the user may be provided a message via an inbox of a first service including the invitation and a selectable option to install an application associated with a second service. In some, examples, the user may be sent a list of contacts from which the user may select other users to also participate in the software testing of the social network application. In some examples, responsive to the user's selection, a message may be sent to an inbox of the selected contacts that contains an invitation to participate in the beta version of the social network application.

Abstract: The present disclosure is directed toward methods, non-transitory computer-readable media, and systems for trajectory optimization in a high-altitude, long-endurance aircraft. For example, the systems described herein can generate an optimized flight plan for an aircraft during active flight of the aircraft by utilizing a greedy algorithm with buffering. In one or more embodiments, the systems described herein identify a plurality of possible states and select a predetermined number of the top possible states (based on energy change associated with transitioning to each possible state) at each incremental time period within a flight time window starting from an initial state to a plurality of possible states for a final incremental time period. Furthermore, in some embodiments, the systems described herein select a final state based on a final energy associated with the final state and determine a flight plan for the aircraft from the initial state to the selected final state.

Abstract: Various client devices include displays and one or more image capture devices configured to capture video data. Different users of an online system are associated with client devices that exchange information captured by their respective image capture devices. When exchanging information, presentation of content to users associated with different client device may be initially synchronized across the client devices. To synchronize content presentation, a client device initiating presentation of the content transmits a request identifying the content and an initial time to other client devices. The initial time is greater than a maximum return time or latency in a network coupling the client devices and the online system from a time when the request is transmitted. A client device determined to be out of synchronization with one or more other client devices receives a command to modify a rate at which the content is presented to reestablish synchronization.

Abstract: In one embodiment, a method includes a computer machine reviewing a picture image and an associated video with the associated video including several video frames. The computer machine then statistically analyzes the corresponding areas for each video frame. The computer machine then compares an area in the picture image with an associated area in a video frame image, wherein the video frame image is based in part on the statistical analysis of the corresponding area for each video frame. Next, the computer machine creates an area for a composite image based in part on the comparison of the area in the picture image with the associated area in the video frame image.

Abstract: Systems, methods, and non-transitory computer-readable media can determine at least one scenario that applies to a user of a social networking system based at least in part on features associated with the user. One or more groups of content recommendations associated with the at least one scenario can be determined. Each group of content recommendations can include a set of content items that relate to the at least one scenario. The one or more groups of content recommendations can be provided to the user as recommendations.

Abstract: Systems, methods, and non-transitory computer-readable media can determine a request for a content item from a computing device. One or more view directions of a user operating the computing device can be predicted during playback of the content item. One or more frames of the content item are quantized based at least in part on the predicted view directions.

Abstract: An online system may identify content with which a user has an interest. For example, the online system may determine that a user has an interest in the content based on interaction information indicating that the user interacted with the content. In a particular example, the online system may identify image concepts included in the content based on computer vision techniques that recognize the image concepts. The online system may model probabilities that image concepts will appeal to users. Based on the modeled probabilities, the online system may automatically recommend image concepts for inclusion in candidate images, automatically generate candidate images, or assess candidate images to determine a probability of user interaction with the assessed candidate images.