Abstract: In one embodiment, a method includes receiving a first image depicting a context including one or more persons having one or more respective poses, receiving a second image depicting a target person having an original pose, where the target person is to be inserted into the context depicted in the first image, generating a target segmentation mask specifying a new pose for the target person in the context of the first image based on the first image, generating a third image depicting the target person having the new pose based on the second image and the target segmentation mask, and generating an output image based on the first image and the third image, the output image depicting the one or more persons having the one or more respective poses and the target person having the new pose.

Abstract: The disclosed apparatus may include unmanned aerial vehicle having (1) a flight system that causes movement and hovering of the unmanned aerial vehicle, (2) a coupling mechanism that interacts with a corresponding coupling mechanism of an electronic device for carrying the electronic device from a remote location to a mount of an infrastructure component, and (3) an extension mechanism connecting the coupling mechanism to the flight system, wherein the extension mechanism dynamically extends the coupling mechanism from the flight system to facilitate installation of the electronic device to the mount while the unmanned aerial vehicle hovers. Various other apparatuses, devices, and methods are also disclosed.

Abstract: An online system receives information including a description of an event occurring in a mobile application and user identifying information associated with a user of the mobile application associated with the event via an SDK incorporated into the mobile application code of the mobile application. The online system determines whether the description corresponds to information maintained in the online system describing types of events capable of occurring in the mobile application. If the description corresponds to information describing a type of event capable of occurring in the mobile application, the online system logs the type of event in association with the user identifying information. Otherwise, the online system predicts a type of event corresponding to the event occurring in the mobile application based at least in part on the information received at the online system and logs the predicted type of event in association with the user identifying information.

Abstract: A social networking system performs account recovery for a user with the help of the user's connections (e.g., friends). The social networking system selects connections of the user based on information indicating likelihood of real-world interactions between the user and the selected connections. Access codes are sent to the selected connections and the user instructed to obtain access codes from the selected connections via a communication that is outside the social networking system, for example, via phone. The user provides the access codes obtained from the selected connections to the social networking system. If the access codes provided by the user match the access codes sent to the selected connections, the user is granted access to the account. Real-world interactions between two users are determined based on sharing of devices between the users or information indicating presence of the users in the same place during same time interval.

Abstract: A geographical area, which forms an area to be represented by a model of at least a portion of a scenery, is defined. A plurality of new images depicting the scenery is received. The plurality of new images is acquired by a camera from one or more positions within the geographical area. It is determined whether the plurality of new images fulfill a pre-defined updating criterion. In response to a determination that the plurality of new images fulfills the updating criterion, an update of the model is initiated. A boundary area enclosing the plurality of new images is defined. The boundary area is defined based at least in part on a distance margin between an edge of the boundary area and a new image included in the plurality of new images that is closest to the edge of the boundary area.

Abstract: The disclosed computer-implemented method may include discovering, by a computer processor and based on user information, that an identity of a user generating a live broadcast is enabled for dual streaming. The method may also include determining, by the computer processor and based at least in part on available network bandwidth, that the live broadcast is authorized for dual streaming. The method may further include transmitting separately, by the computer processor at least partly in response to the discovery and the determination, streams of the live broadcast to a plurality of different network nodes of a streaming platform. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The disclosed computer-implemented method may include systems for optimizing a network environment that is synchronized with a precise time source. For example, a disclosed system can increase the accuracy and efficiency of the network environment with a method for smoothly handing off synchronization control within a group of time servers, each projecting a precise time. Additionally, another disclosed system can further increase the accuracy and efficiency of the network environment with a method for optimizing the latencies of the network environment when scheduling and routing tasks among the network environment members. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Exemplary embodiments relate to methods, mediums, and systems for providing group bots that participate in a conversation between multiple participants. Unlike conventional bots that typically interact with users on a one-on-one basis, the group bots may interact collectively with a group, allowing the bots to (among other things) arbitrate decisions for the group, suggest events that all of the members of the group would find agreeable, facilitate conversations by recognizing core language differences between group members and providing translations in response, etc. Furthermore, the group bot may be invoked directly from a group conversation using natural language, and may provide a response directly in the conversation. Because these capabilities are provided directly within a conversation thread, conversation participants can more readily use the bot capabilities without the need to access separate functionality.

Abstract: According to examples, a system for aligning a plurality of variously encoded content streams is described. The system may include a processor and a memory storing instructions. The processor, when executing the instructions, may cause the system to receive a request to transmit one or more packets of a message over a network, upon receive the request, designate a message sequence number (MSN) for the message, and upon receive the request, designate a packet sequence number (PSN) for one or more packets of the message.

Abstract: In some embodiments, a method includes detecting an incorporation attribute of audiovisual content; analyzing an audio library to determine an audio file that maps to the incorporation attribute; selecting the audio file from the audio library that maps to the incorporation attribute; incorporating the selected audio file into the audiovisual content to generate egocentric audiovisual content; and providing the egocentric audiovisual content to a user for audiovisual consumption. In some embodiments of the method, the incorporation attribute is at least one of a mood of a target of the audiovisual content, an activity of the target of the audiovisual content, and a background of the audiovisual content.

Abstract: In one embodiment, a method includes receiving a user query inputted on a head-mounted device from the head-mounted device, wherein the user query corresponds to multiple dialog-intents, executing multiple tasks corresponding to the multiple dialog-intents, generating a multi-perspective response by a stitching model based on two or more of execution results of the multiple tasks, wherein the stitching model combines the two or more of the execution results based on natural language processing, and wherein the multi-perspective response comprises a natural-language response combining the two or more execution results, and sending instructions to the head-mounted device for presenting the multi-perspective response on the head-mounted device.

Abstract: Systems and methods for creating multiple renditions of a social media composition may include (1) providing a composition screen configured to be used to create user compositions for use with different consumption channels of a social media application, (2) collecting, via the composition screen, a set of one or more user inputs selecting user-selected content, to be included within a user composition, that includes photographs and text, (3) creating a first rendering of the user composition, to be viewed from a first consumption channel, that includes a slideshow of the photographs with a cover card that includes the text, and (4) creating a second rendering of the user composition, to be viewed from a second consumption channel, that includes a collage of the photographs with a caption that includes the text. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The present disclosure is directed to a networking system that provides a user experience in which users of the networking system can interact with user-generated digital content items to select and receive information about one or more products. In particular, the systems and methods involve identifying products shown within user-generated digital content items and associating the user-generated digital content items with the identified products. In addition, where a user selects a product shown within a digital content item, the systems and methods involve selectively identifying other user-generated digital content items for the user based on a coefficient between the user and the other user-generated digital content items. Moreover, the systems and methods involve tracking user-interactions with respect to digital content items and generating and providing a creation insight including an identification of one or more products predicted to engage users of the networking system.

Abstract: The disclosed computer-implemented method may include collecting a set of labels that label polygons within a training set of images as architectural structures. The method may also include creating a set of noisy labels with a predetermined degree of noise by distorting boundaries of a number of the polygons within the training set of images. Additionally, the method may include simultaneously training two neural networks by applying a co-teaching method to learn from the set of noisy labels. The method may also include extracting a preferential list of training data based on the two trained neural networks. Furthermore, the method may include training a machine learning model with the preferential list of training data. Finally, the method may include identifying one or more building footprints in a target image using the trained machine learning model. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A device includes a matrix transpose component, a matrix processing component, a data alignment component, and a data reduction component. The matrix transpose component is configured to transpose an input matrix of elements to output an output matrix of the elements that have been transposed. The matrix processing component is configured to multiply a first multiplication input matrix with a second multiplication input matrix, wherein the output matrix of the matrix transpose component is utilized as the first multiplication input matrix and a mask vector is utilized as the second multiplication input matrix. The data alignment component is configured to modify at least a portion of elements of a result of the matrix processing component. The data reduction component is configured to sum at least the elements of the modified result of the matrix processing component to determine a sum of the group of values.

Abstract: A mixed reality (MR) simulation system includes a console and a head mounted device (HMD). The MR system captures stereoscopic images from a real-world environment using outward-facing stereoscopic cameras mounted to the HMD. The MR system preprocesses the stereoscopic images to maximize contrast and then extracts a set of features from those images, including edges or corners, among others. For each feature, the MR system generates one or more two-dimensional (2D) polylines. Then, the MR system triangulates between 2D polylines found in right side images and corresponding 2D polylines found in left side images to generate a set of 3D polylines. The MR system interpolates between 3D vertices included in the 3D polylines or extrapolates additional 3D vertices, thereby generating a geometric reconstruction of the real-world environment. The MR system may map textures derived from the real-world environment onto the geometric representation faster than the geometric reconstruction is updated.

Abstract: The disclosed computer-implemented method may include receiving an input from a user identifying specified persons that are to be included in a concealed list of persons. The specified persons may be unaware of their inclusion in the concealed list of persons. The method may also include generating the concealed list of persons and determining that a potential participant from the user's concealed list of persons has selected the user in a corresponding concealed list of persons that belongs to the potential participant and, in response to the determination, the method may include creating a private conversation instance that is unique to the user and the potential participant. The generated private conversation instance may include an interaction interface that allows the user and the potential participant to interact in a confidential manner. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Techniques for communicating entity references in messaging groups are described. In one embodiment, an apparatus may comprise a client messaging component operative to receive a message at a recipient client device from a messaging system, the recipient client device associated with a recipient user account, the message comprising an entity reference, the message addressed to a message thread; and extract an augmentation element for the entity reference from the message; and a user interface component operative to display the augmentation element in a message thread display for the message thread on the recipient client device. Other embodiments are described and claimed.

Abstract: A scalable hardware accelerator configured to compute video quality metrics is disclosed. In some embodiments, an accelerator for video quality metrics comprises an application-specific integrated circuit that includes a buffer memory configured to store at least a portion of a reference frame of a video and at least a corresponding portion of a distorted frame of a transcoded version of the video and that includes a processing unit configured to receive data from the buffer memory and compute a perception-based video quality metric for the distorted frame with respect to the reference frame.