Abstract: Systems and methods for information retrieval are described. Embodiments generate a dense embedding for each of a plurality of media objects to be searched, generate a sparse embedding for each of the media objects using an encoder that takes the dense embedding as an input, wherein the sparse embedding satisfies a sparsity constraint that is applied to at least one layer of the encoder during training, and perform a search on the plurality of media objects based at least in part on the sparse embedding.

Abstract: A flexible circuit module is disclosed. The flexible circuit module includes a fabric member including a plurality of fibers having non-conductive elements, and conductive elements oriented through the plane of the fabric. Each conductive element includes a first fiber element strand and a first unterminated end opposite the first fiber element strand, and a second fiber element strand having second unterminated end opposite the second fiber element strand. The fabric member further includes a first side with the first fiber element strands extending therefrom and a second side with second fiber element strands extending therefrom. A first hook-and-loop busbar is conductively connected to the first fiber element strands and a second hook-and-loop busbar is conductively connected to the second fiber element strands.

Abstract: Systems and methods include operating a label distribution protocol (LDP) session with a peer node in a network; responsive to a change in graceful restart of the LDP session, sending a notification message to the peer node with a graceful restart capability change indication; responsive to receiving a second notification message from the peer node with graceful restart capability change indication, updating the graceful restart capability for the LDP session; and, responsive to failing to receive the second notification message, leaving the graceful restart capability as it was prior to the change.

Abstract: An illustrator system accesses a multi-element document, the multi-element document including a plurality of elements. The illustrator system determines, for each of the plurality of elements, an element-specific topic distribution comprising a ranked list of topics. The illustrator system creates a first aggregated topic distribution from the determined element-specific topic distributions. The illustrator system determines a global intent for the multi-element document, the global intent including one or more terms from the first aggregated topic distribution. The illustrator system queries a database using the global intent to retrieve a substitute element. The illustrator system generates a replacement multi-element document that includes a substitute element in place of an element in the multi-element document The at least one substitute element is different from the element in the displayed multi-element document.

Abstract: Embodiments are disclosed for receiving a modal input including at least one of a text input or an image input. The method may include extracting an intent label from the modal input. The method may further include generating, by an intent embedding generator, an intent embedding from the intent label. The method may further include comparing the intent embedding to a plurality of candidate font embeddings to obtain one or more candidate fonts based on a similarity of the intent embedding to the plurality of candidate font embeddings in an embedding space. The method may further include identifying a recommended font based on the similarity of the intent embedding to a selected candidate font embedding of the plurality of candidate font embeddings.

Abstract: A method for optical camera quality enhancement is provided. The method includes receiving, by an electronic device, an image frame of a scene including at least one object from a preview field of a camera of the electronic device, determining, by the electronic device, at least one region of the image frame affected by distortion, generating, by the electronic device, a pixel disparity map comprising a refraction coefficient associated with each pixel of the image frame based on the at least one region affected by the distortion, determining, by the electronic device, a plurality of regions on the pixel disparity map, each of the plurality of regions including pixels that have a similar pixel disparity, estimating, by the electronic device, an optimal orientation of the electronic device to minimize the distortion in the image frame based on the plurality of regions on the pixel disparity map, and displaying, by the electronic device, an indication indicating the optimal orientation of the electronic device.

Abstract: Systems and methods for image processing are described. Embodiments of the present disclosure receive an image depicting an object; generate a sequence of tokens including a set of tokens corresponding to the object and a set of mask tokens corresponding to an additional object to be inserted into the image; generate a placement token value for the set of mask tokens based on the sequence of tokens using a sequence encoder, wherein the placement token value represents position information of the additional object; and insert the additional object into the image based on the position information to obtain a composite image.

Abstract: Techniques are disclosed in which a server computer system manages a database cache for multiple different tenants. The system may retrieve, from a database cache storing data for different tenants, a set of key-value entries and cache metadata for the set of key-value entries. Further, the system may identify, based on analyzing keys of key-value entries, tenants corresponding to respective ones of the set of key-value entries. The system may generate, based on the identified tenants and cache metadata for cache entries corresponding to the identified tenants, current usage metrics for the database cache. Based on the current usage metrics for the cache, the system may determine whether to adjust one or more operation parameters of the cache. Disclosed cache management techniques may advantageously allow for analysis of cache usage metrics for individual tenants and more efficient use of the cache (e.g., by altering cache parameters for individual tenants).

Abstract: An illustrator system accesses a multi-element document, the multi-element document including a plurality of elements. The illustrator system determines, for each of the plurality of elements, an element-specific topic distribution comprising a ranked list of topics. The illustrator system creates a first aggregated topic distribution from the determined element-specific topic distributions. The illustrator system determines a global intent for the multi-element document, the global intent including one or more terms from the first aggregated topic distribution. The illustrator system queries a database using the global intent to retrieve a substitute element. The illustrator system generates a replacement multi-element document that includes a substitute element in place of an element in the multi-element document The at least one substitute element is different from the element in the displayed multi-element document.

Abstract: Disclosed are systems and methods for dynamically determining categories for images. A computer-implemented method may include training a neural network to receive an input image and determine one or more image categories associated with the input image; obtaining a set of images associated with a user; determining, using the trained neural network, one or more image categories associated with each image included in the obtained set of images; determining one or more dominant image categories associated with the user based on the determined image categories for the obtained set of images; and determining an image editing user interface for the user based on the determined one or more dominant image categories.

Abstract: A Point of Local Repair (PLR) network element includes one or more ports and circuitry connected thereto for forwarding and control, wherein the circuitry is configured to receive a PATH message for a Label Switched Path (LSP) tunnel in a Multiprotocol Label Switching (MPLS) network with a specified DiffSery Traffic Engineering (DSTE) Class Type, determine the DSTE Class Type based on the PATH message, and store the DSTE Class Type for the LSP tunnel to ensure a Facility Bypass tunnel used for the LSP tunnel supports the specified DSTE Class Type. The circuitry can be further configured to, responsive to a failure of the LSP tunnel, select the Facility Bypass tunnel for the LSP tunnel such that the Facility Bypass tunnel supports the specified DSTE Class Type.

Abstract: This disclosure describes methods, non-transitory computer readable storage media, and systems that generate a digital knowledge graph based on a plurality of tutorial content items to generate recommendations of digital resource items. Specifically, the disclosed system extracts a plurality of tasks, subject categories related to the tasks, and context signals related to an environment for the tasks from a plurality of tutorial content items for one or more digital content editing applications. The disclosed system generates a digital knowledge graph including nodes corresponding to the tasks and subject categories connected via edges based on relationships extracted from the tutorial content items. In some embodiments, the disclosed system also includes nodes corresponding to digital resource items in the digital knowledge graph or in a subgraph. The disclosed system utilizes the digital knowledge graph with context data to provide a recommendation of digital resource items for display at a client device.

Abstract: Techniques are disclosed in which a server computer system manages a database cache for multiple different tenants. The system may retrieve, from a database cache storing data for different tenants, a set of key-value entries and cache metadata for the set of key-value entries. Further, the system may identify, based on analyzing keys of key-value entries, tenants corresponding to respective ones of the set of key-value entries. The system may generate, based on the identified tenants and cache metadata for cache entries corresponding to the identified tenants, current usage metrics for the database cache. Based on the current usage metrics for the cache, the system may determine whether to adjust one or more operation parameters of the cache. Disclosed cache management techniques may advantageously allow for analysis of cache usage metrics for individual tenants and more efficient use of the cache (e.g., by altering cache parameters for individual tenants).

Abstract: A node, in a first network, includes circuitry configured to determine a next hop as decided by Border Gateway Protocol (BGP) is an anycast prefix to a Route Reflector (RR) interconnecting the first network with a second network, responsive to the next hop being the anycast prefix to the RR, create a tunnel with a destination based on the anycast prefix, and utilize the tunnel for traffic having the next hop as the anycast prefix to the RR. The anycast prefix is assigned to two or more RRs interconnecting the first network and the second network. A first path is decided by BGP based on a BGP Path Selection Algorithm that is independent of a second path determined by Interior Gateway Protocol (IGP). The first path and the second path can be different, and wherein tunnel is utilized to ensure the traffic always follows the first path.

Abstract: This disclosure describes methods, non-transitory computer readable storage media, and systems that generate a digital knowledge graph based on a plurality of tutorial content items to generate recommendations of digital resource items. Specifically, the disclosed system extracts a plurality of tasks, subject categories related to the tasks, and context signals related to an environment for the tasks from a plurality of tutorial content items for one or more digital content editing applications. The disclosed system generates a digital knowledge graph including nodes corresponding to the tasks and subject categories connected via edges based on relationships extracted from the tutorial content items. In some embodiments, the disclosed system also includes nodes corresponding to digital resource items in the digital knowledge graph or in a subgraph. The disclosed system utilizes the digital knowledge graph with context data to provide a recommendation of digital resource items for display at a client device.

Abstract: A system and method for automatically detecting a location of a plurality of hotspots from a thermal image of a breast region of a subject by (i) automatically detecting areolar points (x, y) from the thermal image of the breast region of the subject, (ii) automatically detecting a plurality of hotspot regions on the thermal image of the breast region of the subject by performing a hotspot region segmentation method, (iii) calculating a plurality of radial locations (ri, ?) of a plurality of hotspots on the hotspot region, (iv) automatically generating a text report based on the detected location of the plurality of hotspots and (v) providing the detected radial locations (ri, ?i) of the plurality of hotspots as a text report to scan the plurality of hotspots only on the detected radial locations (ri, ?i) instead of scanning in the entire breast region of the subject.

Abstract: The present disclosure relates to an image merging system that automatically and seamlessly detects and merges missing people for a set of digital images into a composite group photo. For instance, the image merging system utilizes a number of models and operations to automatically analyze multiple digital images to identify a missing person from a base image, segment the missing person from the second image, and generate a composite group photo by merging the segmented image of the missing person into the base image. In this manner, the image merging system automatically creates merged group photos that appear natural and realistic.

Abstract: A node, in a first network, includes circuitry configured to add a label stack to a packet associated with one of a Layer 2 Virtual Private Network (L2VPN) and a Layer 3 Virtual Private Network (L3VPN), wherein the packet is destined for a second network having at least a pair of Border Gateway Protocol-Label Unicast (BGP-LU) Route Reflector (RR) nodes interconnecting the first network, and wherein the label stack includes an anycast Segment Identifier (SID) for the RR nodes, in Segment Routing, and transmit the packet to the second network. The label stack also includes a BGP-LU label for a destination in the second network. Responsive to isolation of one of the RR nodes, a Point of Local Repair (PLR) node in the first network is configured to implement Topology-Independent Loop-Free Alternate (TI-LFA) for the anycast SID.

Abstract: A node, in a first network, includes circuitry configured to add a label stack to a packet associated with one of a Layer 2 Virtual Private Network (L2VPN) and a Layer 3 Virtual Private Network (L3VPN), wherein the packet is destined for a second network having at least a pair of Border Gateway Protocol-Label Unicast (BGP-LU) Route Reflector (RR) nodes interconnecting the first network, and wherein the label stack includes an anycast Segment Identifier (SID) for the RR nodes, in Segment Routing, and transmit the packet to the second network. The label stack also includes a BGP-LU label for a destination in the second network. Responsive to isolation of one of the RR nodes, a Point of Local Repair (PLR) node in the first network is configured to implement Topology-Independent Loop-Free Alternate (TI-LFA) for the anycast SID.

Abstract: Disclosed are systems and methods for dynamically determining categories for images. A computer-implemented method may include training a neural network to receive an input image and determine one or more image categories associated with the input image; obtaining a set of images associated with a user; determining, using the trained neural network, one or more image categories associated with each image included in the obtained set of images; determining one or more dominant image categories associated with the user based on the determined image categories for the obtained set of images; and determining an image editing user interface for the user based on the determined one or more dominant image categories.