Abstract: The disclosed computer-implemented method may include (i) identifying a set of resources stored on a server that are requested by a browser when the browser loads a web page, (ii) determining a first modification pattern with which a first resource within the set of resources is modified, (iii) determining a second modification pattern with which a second resource within the set of resources is modified, (iv) evaluating a compatibility between the first modification pattern and the second modification pattern, and (v) packaging, based at least in part on the compatibility between the first modification pattern and the second modification pattern, the first resource and the second resource into a package that is sent by the server to the browser via a single network protocol request. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A method comprises inputting an audio signal into a machine learning circuit to compress the audio signal into a sequence of actuator signals. The machine learning circuit being trained by: receiving a training set of acoustic signals and pre-processing the training set of acoustic signals into pre-processed audio data. The pre-processed audio data including at least a spectrogram. The training further includes training the machine learning circuit using the pre-processed audio data. The neural network has a cost function based on a reconstruction error and a plurality of constraints. The machine learning circuit generates a sequence of haptic cues corresponding to the audio input. The sequence of haptic cues is transmitted to a plurality of cutaneous actuators to generate a sequence of haptic outputs.

Abstract: Systems, methods, and non-transitory computer-readable media can determine a set of candidate values for a field in a page. The set of candidate values can be evaluated for accuracy based at least in part on a machine learning model, wherein the machine learning model outputs a respective score for each candidate value that measures an accuracy of the candidate value for the field in the page. A best scoring candidate value can be determined from the set of candidate values. The field in the page can be associated with the best scoring candidate value.

Abstract: Systems and methods for creating and maintaining digital story progress bars may include (1) providing an interface that displays posts within a series of different digital stories, (2) displaying, within the interface, a progress bar including (i) a current-story section associated with a story-indicator corresponding to a current story whose posts are currently being displayed, and (ii) an upcoming-story section associated with a subsequent story-indicator corresponding to an upcoming story queued for display after the current story, (3) transitioning from displaying the posts of the current story to displaying the posts of the upcoming story, and (4) in response to displaying the posts of the upcoming story (i) altering the current-story section by replacing the story-indicator with the subsequent story-indicator and (ii) altering the upcoming-story section by replacing the subsequent story-indicator with a post-subsequent story-indicator corresponding to an additional upcoming story.

Abstract: Techniques are described that classify content, and control whether and how the content is shared based on the classification(s). In some examples, video content may be classified based on sequential image frames of the video, and time between the sequential image frames. Audio content may be classified based on combining classifications of multiple sound events in the audio signal. The classifications may be used to control how the content is shared, such as by preventing offensive content from being shared and/or outputting recommendations or search results based on the classifications.

Abstract: Systems, methods, and non-transitory computer-readable media can receive from a user computing device raw stack trace information associated with an exception that has occurred within a computing application on the user computing device. Debugging information associated with the computing application is retrieved. Human-readable stack trace information is generated based on the raw stack trace information and the debugging information.

Abstract: Systems, methods, and non-transitory computer-readable media can provide an interface in which a content feed of a user is presented, the content feed including at least one first post that was authored by the user and that includes a first content item, wherein the interface is presented on a display screen of a computing device. A determination is made that the user has selected an option to update the first post to include one or more additional content items. A set of content items is provided for selection through the interface. The first post is updated to include at least one content item that was selected by the user.

Abstract: Systems, methods, and non-transitory computer-readable media can group a set of live content items based at least in part on the live content items satisfying at least one relatedness criteria, the live content items being broadcasted through the content provider system. The set of live content items is ranked. An interface that includes a content feed through which the ranked set of live content items are accessible is generated.

Abstract: A computing system is configured to receive location data from a wireless device of a user. Upon receiving a search request from the wireless device, the computing system utilizes the location data to determine one or more likely travel routes of the user of the wireless device. In response to the search request, the computing system may select one or more establishments based on the likely travel routes and location data associated with the one or more establishments. The computing system may further transmit to the wireless device information associated with the selected one or more establishments.

Abstract: An online system allows users to exchange messages with other users, including entities such as businesses. A user, such as a business, may use a third party system to manage messages with users of the online system. The online system provides the third party information with instructions comprising tracking mechanisms that identify specific actions between users from messages exchanged between users. For example, if a user purchases a product from another user via a message thread, a tracking mechanism identifies the purchase and transmits information reporting the purchase and identifying users to the online system. Hence, the tracking mechanism transmits a description of an action and information identifying the users to the online system, and may also transmit additional information, such as a value of the action to a user. The online system may use the descriptions of actions for content targeting or measurement and reporting of actions.

Abstract: A communication system captures writing surface content in a physical space for transmittal to remote client devices participating in a communication session. During a communication session with one or more remote client devices, a communication system captures image data and depth data describing objects in a physical space of the communication system. Based on the captured data, the communication system identifies a writing surface in the physical space and captures content on the writing surface. The communication system may also identify objects occluding content on the writing surface based on the captured data and may modify image data to make an object occluding the content at least partially transparent. The communication system transmits the content to at least one of the remote client devices participating in the communication session.

Abstract: A messaging system receives a message sent by an enterprise to an individual user. The message has an associated message tag describing the content of the message. The messaging system applies a filtering policy to the message. The filtering policy selectively blocks messages sent by enterprises to users. The filtering policy allows the message having the associated tag to pass through the filter even though the message might otherwise violate the filtering policy. The messaging system samples a subset of tagged messages from enterprises and analyzes the messages for compliance with a tagging policy. The messaging system may also train one or more tag models to recognize the correct tags for the messages.

Abstract: A computer-implemented method for digital content navigation based on directional input may include (i) maintaining a user content pod for each user account within a group of user accounts maintained by a social networking platform, (ii) generating a cover card for each user content pod, and (iii) providing, within a graphical user interface for interacting with posts made to the social networking platform, a scrollable display of vertically arranged selectable cover cards configured to (1) scroll vertically between cover cards of different user content pods in response to receiving vertical swiping input and (2) scroll horizontally between one or more posts of an individual user content pod in response to receiving horizontal swiping input. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Machine-trained models are generated based on a model description that defines parameters for training the model and that can inherit parameters from parent model descriptions. When a parent model description changes, the changes made to the parent model description are applied to the model description automatically. When a target model is re-generated, a description of the set of parameters for generating the target model is received. The parent model is then identified from the received description, and a description of the set of parameters for generating the parent model is retrieved. Using the description for the target model and the parent model, a pipeline for generating the target model is generated. Finally, the pipeline is executed to generate the target model.

Abstract: Techniques for a system capable of performing low-latency database query processing are disclosed herein. The system includes a gateway server and a plurality of worker nodes. The gateway server is configured to divide a database query, for a database containing data stored in a distributed storage cluster having a plurality of data nodes, into a plurality of partial queries and construct a query result based on a plurality of intermediate results. Each worker node of the plurality of worker nodes is configured to process a respective partial query of the plurality of partial queries by scanning data related to the respective partial query that stored on at least one data node of the distributed storage cluster and generate an intermediate result of the plurality of intermediate results that is stored in a memory of that worker node.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods for improving and streamlining user engagement with co-users within a networking system. For instance, the user engagement system can detect co-user engagement with a user of a networking system. Based on one or more metrics, such as characteristics of the engaging co-user, the type of detected co-user engagement, or engagement history with the user, the user engagement system can rank, prioritize, and/or aggregate the engagement notifications. For example, the user engagement system can prioritize message requests from co-users with whom the user has not actively engaged on the networking system. In another example, the user engagement system can filter messages from co-users with whom the user has actively engaged in to one of multiple message repositories.

Abstract: The disclosed computer-implemented method may include receiving original media content, transcoding the original media content into a plurality of encoded media files, determining an adaptive bit rate for providing a one of the plurality of encoded media files to a computing device, delivering the encoded media file to the computing device at the adaptive bit rate for viewing on a display device of the computing device, and measuring a quality of the displayed encoded media file based on an overall mean opinion score associated with the delivered encoded media file. The overall mean opinion score may be based on a measured playback quality associated with the encoded media file and a measured upload quality associated with the original media content. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Systems and methods are disclosed herein for activating a sensor of a client device by toggling, at the client device, a position of a switch that connects the sensor to a power source of the client device to an on position. The client device receives user input corresponding to a privacy mode. In response to receiving the user input, the client device disables the sensor by toggling the position of the switch that connects the sensor to the power source to an off position, and enables a privacy indicator by toggling a position of a switch that connects the privacy indicator to a power source to an on position.

Abstract: A robotic-arm apparatus may include a robotic hand dimensioned to approximate a size and movement of a user's hand. The robotic-arm apparatus may also include one or more tactile-sensing pads coupled to at least a portion of the robotic hand, wherein a tactile-sensing pad is configured to detect surface data about a surface in a real-world environment. Additionally, the robotic-arm apparatus may include an actuator configured to move the robotic hand to mimic a motion of a glove worn by the user's hand, wherein the glove is configured to provide haptic feedback corresponding to the surface data to the user's hand. Various other apparatuses, systems, and methods are also disclosed.

Abstract: A disclosed computer-implemented method may include (1) receiving, from a participant of a discovery service of a social networking system, a request for a friend of the participant to assist the participant in making connections with other participants of the discovery service, (2) identifying at least one set of information that is associated with the discovery service and that is designated as private to the participant, and (3) providing the friend of the participant with access to at least a portion of the set of information via a recommendation interface that enables the friend of the participant to make recommendations to the participant regarding the set of information to assist the participant in making connections with other participants of the discovery service. Various other methods, systems, and computer-readable media are also disclosed.