Abstract: Technology for identifying an object of interest includes obtaining object embeddings for a plurality of objects in an image, obtaining text embeddings for text associated with the image, determining, for each of the plurality of objects, a similarity score via a similarity model based on the text embeddings and the object embeddings, while bypassing use of bounding box coordinates, and selecting the object having the highest similarity score as the object of interest. In another example, technology for identifying an object of interest includes obtaining object embeddings for a plurality of objects in an image, obtaining text embeddings and text identifiers for text associated with the image, generating, via a single transformer encoder, a set of CLS embeddings based on the text embeddings and the object embeddings, and determining, via a neural network, the object of interest based on the CLS embeddings.

Abstract: Systems, methods, and non-transitory computer-readable media can be configured to determine a life story associated with a first user of a content sharing system. The life story can comprise a plurality of content items associated with the first user. The plurality of content items can be organized in chronological order. A determination is made that a second user has accessed the life story associated with the first user. The plurality of content items included in the life story are provided for presentation in a graphical user interface, wherein each content item is presented individually based on the chronological order.

Abstract: Techniques enabling accounts of a social networking system to monitor online interactions by supervised accounts across multiple platforms are described. For example, a social networking system (SNS) may receive, from a first account, a request to invite a second account to join a Family Center (FC) so the FC includes data associated with first and second platforms of the SNS. The SNS may receive, from the second account, an indication to accept the request to join the FC associating the first and second accounts via the FC. In some examples, the SNS may receive, from the second account, a request to access the FC. Based on receiving the request to access the FC, the SNS may cause presentation to the second account of interaction data associated with interactions between the first account and the first platform of the SNS and the first account and the second platform of the SNS.

Abstract: Techniques for notifying parent accounts of problematic interactions by teens is described. For example, a social networking system may receive, from a first account, an indication of a problematic interaction. Based at least in part on receiving the indication, the social networking system may cause presentation, to the first account, of a prompt to notify a second account associated with the first account of the interaction. Based at least in part on receiving a confirmation from the first account, the social networking system may then send, to the second account, a notification of the problematic interaction.

Abstract: Techniques for notifying parent accounts of problematic interactions by teens is described. For example, a social networking system may receive, from a first account of the social networking system, an indication of an interaction between the first account and a content item. Based on determining that the interaction is a problematic interaction, the social networking system may send, to a second account associated with the first account, a notification of the problematic interaction. The social networking system may cause presentation, to the second account, of a selectable control associated with one or more resources regarding addressing the problematic interaction.

Abstract: A computer-implemented method for client-side identity verification may include (1) receiving, via an endpoint computing device, input from a user that includes biometric data of the user captured by a sensor of the endpoint computing device and visual data of a physical identification document that includes a record of the biometric data, (2) verifying, by the endpoint computing device, that the biometric data captured by the sensor of the endpoint computing device matches the record of the biometric data in the physical identification document, and (3) transmitting, to a server, a verification that the user has been identified while preventing the biometric data from being included in the verification sent to the server. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: In one embodiment, a method for machine learning acceleration includes receiving instructions to perform convolution on an input tensor using a filter tensor, determining that the size of a first dimension of the input tensor is less than a processing capacity of each of multiple subarrays of computation units in a tensor processor, selecting a second dimension of the input tensor along which to perform the convolution, selecting, based on the second dimension, one or more dimensions of the filter tensor, generating (1) first instructions for reading, using vector read operations, activation elements in the input tensor organized such that elements with different values in the second dimension are stored contiguously in memory, and (2) second instructions for reading weights of the filter tensor along the selected one or more dimensions, and using the first and second instructions to provide the activation elements and the weights to the subarrays.

Abstract: A speech translation system and methods for cross-lingual communication that enable users to improve and customize content and usage of the system and easily. The methods include, in response to receiving an utterance including a first term associated with a field, translating the utterance into a second language. In response to receiving an indication to add the first term associated with the field to a first recognition lexicon, adding the first term associated with the field and the determined translation to a first machine translation module and to a shared database for a community associated with the field of the first term associated with the field, wherein the first term associated with the field added to the shared database is accessible by the community.

Abstract: In one embodiment, a method includes receiving, at a client system, an audio input, where the audio input comprises a coreference to a target object, accessing visual data from one or more camera associated with the client system, where the visual data comprises images portraying one or more objects, resolving the coreference to the target object from among the one or more objects, resoling the target object to a specific entity, and providing, at the client system, a response to the audio input, where the response comprises information about the specific entity.

Abstract: In one embodiment, a method includes rendering a first output image comprising one or more augmented-reality (AR) objects for displays of an AR rendering device of an AR system associated with a first user. The method further includes accessing sensor signals associated with the first user. The one or more sensor signals may be captured by sensors of the AR system. The method further includes detecting a change in a context of the first user with respect to a real-world environment based on the sensor signals. The method further includes rendering a second output image comprising the AR objects for the displays of the AR rendering device. One or more of the AR objects may be adapted based on the detected change in the context of the first user.

Abstract: Ephemeral stories allow users of an online system, such as a social networking system, to share content for a pre-determined temporary length of time. Channels group and organize stories based on content. Topic-based channels allow contain stories that have related content. Hashtag- and sticker-associated channels contain stories that include particular hashtags or stickers, respectively. User-specific channels contain content tailored by the online system for display to a particular user based on preferences and history. Collaborative channels are channels that are private to a specific group of users. The use of channels improves the online experience by grouping stories by content, thereby facilitating users' interaction with similar content.

Abstract: A method for generating a content feed includes receiving, from a content provider, a target content item and a policy for the target content item, specifying a prohibited classification of the content that cannot be published adjacent to the target content item. The method also includes identifying a slot in a content feed including multiple content items. A machine learning model is then accessed and applied to two adjacent content items that are adjacent to the slot to determine whether each of the adjacent content items is the prohibited classification. Responsive to determining that the adjacent content items are not the prohibited classification, the target content item is placed in the slot, and the content feed including the target content item is sent for display to a viewing user.

Abstract: In one embodiment, a method for machine learning acceleration includes receiving, by a shared controller of a tensor processor cluster that includes multiple tensor processors, a multi-cycle instruction, determining, based on the instruction, a sequence of vector operations to be executed by the tensor processors and address information usable to determine a respective spatial partition of an input tensor on which each tensor processor is to operate when performing each vector operation. The method also includes, for each vector operation in the sequence, generating, based on the address information, a common address offset, relative to a respective base address associated with each tensor processor, at which each tensor processor is to retrieve the respective spatial partition on which the tensor processor is to operate, multicasting the common address offset to the tensor processors, and controlling the tensor processors to execute the vector operation in parallel and in lock step.

Abstract: Disclosed technology herein provides a computer-implemented method for asset tracking in a network having a plurality of connected devices, comprising generating a network graph based on network data captured at least in part via peer-to-peer polling, the network data including data relating to tracked assets in a network, where the peer-to-peer polling is used to validate a presence of at least a portion of the tracked assets in the network, generating an estimated size of the network graph using a graph estimation algorithm, determining an estimate of the tracked assets based on the estimated size of the network graph, and determining one or more remediation actions in response to the estimate of the tracked assets, where at least a portion of the network graph can be stored in a decentralized manner. The graph estimation algorithm can include a random walk algorithm, a random sampling algorithm, and/or an induced edges algorithm.

Abstract: Systems and methods for availability-based streaming may include (1) detecting that a user of a messaging service has selected a call-availability element provided within an instance of a messaging thread, generated for the user, that enables digital messaging between the user and a contact of the user, (2) signaling to the user's contact that the user is available for a call via the messaging thread by instructing a device of the contact to display a selectable prompt that indicates that the user is available for the call via the messaging thread, (3) detecting that the user's contact has selected the selectable prompt via the contact's device, and (4) instructing the contact's device to display a call-launch element configured to launch the call between the user and the user's contact via the messaging thread when selected via user input. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: In one embodiment, a method includes displaying, by the first computing device, a first content adaption of a digital content, synchronously with one or more second content adaptions displayed on one or more second computing devices, respectively, where the one or more second computing devices are in physical proximity to the first computing device, detecting, by the first computing device, a change in position information for at least one of the one or more second computing devices, receiving, by the first computing device from a server, a reconfigured first content adaption, where the reconfigured first content adaption and one or more reconfigured second content adaptions are reconfigured responsive to the change in position information, and displaying, by the first computing device, the reconfigured first content adaption, synchronously with one or more of the reconfigured second content adaptions displayed on one or more of the second computing devices, respectively.

Abstract: In one embodiment, a method includes receiving a user request from a client system associated with a user, determining that the user request corresponds to a first suspended task, retrieving a first dialog state of the first suspended task from a dialog history associated with the user, generating a summary of the first suspended task based on the first dialog state using a natural-language generating (NLG) module, and sending instructions to the client system for providing the summary of the first suspended task to the user.