Abstract: Embodiments are directed to system and processes for adding a video object to a content panel of a content collaboration system. The processes can include instantiating a client application displaying a graphical user interface having a content panel operating an editor configured to receive input from the client device and receiving a command to initiate creation of a video object. In response to receiving the command, the processes can invoke a video-creation service overlaying the interface and including a content region, a user video region showing a video preview of a user of the client device and a first selectable option configured to initiate a recording session for the video object. The system can initiate a recording session for the video object and cause display of the recoded video object in an editing interface. The video object can be added to the content panel of the content collaboration system.

Abstract: Example aspects of the present disclosure are directed to systems and methods that learn a compressed representation of a machine-learned model (e.g., neural network) via representation of the model parameters within a reparameterization space during training of the model. In particular, the present disclosure describes an end-to-end model weight compression approach that employs a latent-variable data compression method. The model parameters (e.g., weights and biases) are represented in a “latent” or “reparameterization” space, amounting to a reparameterization. In some implementations, this space can be equipped with a learned probability model, which is used first to impose an entropy penalty on the parameter representation during training, and second to compress the representation using arithmetic coding after training. The proposed approach can thus maximize accuracy and model compressibility jointly, in an end-to-end fashion, with the rate-error trade-off specified by a hyperparameter.

Abstract: A robotic system includes a mobile base assembly and a lifting assembly. The lifting assembly includes wedge blocks, each of which is pivotably connected to a front edge of the mobile base assembly and pivotable along a hinge axis. During an operation of the robotic system, at least a portion of the wedge blocks may be slid under the boxes to lift the boxes.

Abstract: The system can include a set of joints, a controller, and a model engine; and can optionally include a support structure and an end effector. Joints can include: a motor, a transmission mechanism, an input sensor, and an output sensor. The system can enable articulation of the plurality of joints.

Abstract: A system and method are disclosed for aggregating product trends. Embodiments include receiving an initial set of images from one or more data feeds by a trend aggregation system comprising a server, identifying one or more attribute values from the initial set of images, and quantifying a social affinity score of one or more items of an inventory of a supply chain entity based on recentness, relevance, and similarities of the identified one or more attribute values to an attribute value of a potential product for a product assortment. Embodiments may further include receiving an image of at least one additional item, identifying a product attribute from the image, and assigning an attribute value to the at least one additional item based, at least in part, on the identified attribute value from the image of the least one additional item and the attribute values of the at least two items.

Abstract: An apparatus to facilitate tessellation redistribution for reducing latencies in processors is disclosed. The apparatus includes a processor to provide parallel interconnected geometry fixed-function units with separate front end and back ends, the front ends to perform patch culling and transmission and the back ends to perform patch reception from the front end and patch tessellation; provide a tessellation redistribution central engine to redistribute patches among the back ends using a redistribution bus; receive, by the tessellation redistribution central engine from the front ends in parallel, patch transmissions marked for distribution, the tessellation redistribution engine to process the patch transmissions in order; and in response to receiving a synchronization barrier packet from one of the front ends, broadcast, by the tessellation redistribution central engine, the synchronization barrier packet to the back ends to cause one of the back ends to process tessellation work locally.

Abstract: Methods, systems, and storage media for running unified simulations on clusters. Exemplary implementations may include: receiving simulation parameters for a simulation of a cluster; generating synthesized workload events based the simulation parameters of the cluster; determining a memory latency associated with the cluster; determining a reliability and availability of resources in the cluster for a predetermined duration of time; simulating events for jobs in the cluster based on the reliability and availability of resources in the cluster, each job associated with one or more synthesized workload events; and outputting simulation results based on the synthesized workload, the memory latency, and the events.

Abstract: The technology disclosed relates to analysis of security posture of a cloud environment. In particular, the disclosed technology relates to a system and method of risk event detection and remediation. An event is detected in a cloud environment and a pre-defined risk signature is obtained that identifies one or more entities in the cloud environment and represents an instance of a risk event relative to the one or more entities. The pre-defined risk signature includes a reference to a remediation workflow having one or more commands for one or more remediation actions in the cloud environment. Th pre-defined risk signature is determined to have a threshold match to the event and, based on the determination that the pre-defined risk signature has a threshold match to the event, the remediation workflow is obtained based on the reference. The one or more commands are executed in the cloud environment.

Abstract: The technology disclosed relates to a system and method for detecting risk events in cloud environment that obtains set of risk signature definitions and deploys an event log scanner to the cloud environment. The event log scanner is configured to detect instances of candidate risk events in accordance with the set of risk signature definitions based on a scan of event log and to label each detected instance with a signature identifier that identifies one or more risk signatures that corresponds to the detected instance. Result metadata is received indicative of the detected instances, based on the result metadata, context information associated with the detected instances is obtained based on cloud infrastructure graph. An output is generated representing a classification of one or more of the detected instances of candidate risk events as a risk event based on the context information relative to the set of risk signature definitions.

Abstract: Disclosed is a system including a user device, a frame, a drone, and a server. The user device enables a user to aviate the drone that is adapted to be detachably coupled to the frame. The frame is further adapted to be detachably coupled to the one or more powerlines, and includes an X-ray unit configured to capture X-ray images of the one or more powerlines. The X-ray unit transmits the X-ray images to the user device and the user device further transmits the X-ray images to the server in a digital imaging and communication in non-destructive evaluation (DICONDE) format. The server detects one or more abnormalities in the one or more powerlines by processing the X-ray images using a picture archiving & communication system (PACS) and an ADR technique, and generates data and/or reports to be viewed by the user by way of the user device.

Abstract: A content collaboration platform, such as described herein, may include dynamic object tables that are configured to dynamically populate and/or update. Dynamic object tables may be configured to retrieve content (e.g., page content) from the content collaboration platform and automatically populate attributes from that content directly in the cells of the dynamic object tables. Updates to the page content may also be automatically updated within the table. In some examples, the dynamic object tables may be embedded within pages of the content collaboration platform and, within the same page, the user can edit, filter, and sort the dynamic object table without toggling to other tabs, applications, or documents.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed to automate suspension and redeployment of cloud resources. The example apparatus is to, based on network traffic associated with a compute cluster hosting a containerized application, determine whether to suspend the containerized application. Additionally, the example apparatus is to determine a port of a transient container that is available to be mapped to the containerized application and cause a request to access the containerized application to be forwarded to the port of the transient container instead of the containerized application. The example apparatus is also to deprovision one or more resources associated with the containerized application.

Abstract: An intelligent editor for creating a pipeline configuration file. The pipeline configuration file can be a YAML configuration file or in some other data serialization language. The intelligent editor is in the form of a YAML editor that provides multiple user interfaces for editing a pipeline configuration file. The YAML editor includes a code-based editing UI to edit the YAML configuration file code itself, and a visual-based editor UI for creating and modifying the configuration file through using graphical icons. The code-based editor UI includes an intelligent field and value recommendation feature, auto-complete feature, a semantic error detection feature, and a field documentation feature.

Abstract: The disclosed technology teaches reducing threat detection processing by applying similarity measures. The method includes recognizing that a file is an edited version of a previously processed file and retrieving, from an archive, at least an entropy measure of the previously processed file, and calculating an entropy measure for the edited version of the file. The method applies a similarity measure to compare the entropy measures for the edited version and the previously processed file, avoiding full threat scanning of the file to detect malware except when the similarity measure reaches a scanning trigger. When any similarity measure or combination of similarity measures reaches a trigger, the technology teaches processing the file by using a threat detection module to detect malware. Further included is logging the edited version of the file for further processing when the similarity measure reaches a logging trigger.

Abstract: The disclosed technology teaches reducing threat detection processing, including recognizing that a file is an edited version of a previously processed file and retrieving, from an archive, metadata values, hashes for property groups and an entropy measure of the previously processed file. Also included is parsing the file into metadata values and property groups and calculating hashes of the property groups and entropy measure for the file. The method further includes applying similarity measures to compare the metadata values, the entropy measures, and the hashes on the property groups, for the edited version and the previously processed file. When any similarity measure or combination of similarity measures reaches a trigger, the technology teaches processing the file by using a threat detection module to detect malware. Property groups include core properties, application properties, document content and programming scripts for the edited version of the file and the previously processed file.

Abstract: The disclosed technology teaches reducing threat detection processing by applying similarity measures. The method includes recognizing that a file is an edited version of a previously processed file and retrieving, from an archive, at least an entropy measure of the previously processed file, and calculating an entropy measure for the edited version of the file. The method applies a similarity measure to compare the entropy measures for the edited version and the previously processed file, avoiding full threat scanning of the file to detect malware except when the similarity measure reaches a scanning trigger. When any similarity measure or combination of similarity measures reaches a trigger, the technology teaches processing the file by using a threat detection module to detect malware. Further included is logging the edited version of the file for further processing when the similarity measure reaches a logging trigger.

Abstract: Techniques for optimizing network-transferred multi-card content items are provided. In one technique, a first content item selection event is initiated that involves a set of content delivery campaigns that includes a content delivery campaign that includes a content item that comprises multiple cards. The content delivery campaign is selected and the content item is transmitted to a first computing device, where the multiple cards have a first card configuration. One or more events that are associated with the first computing device displaying at least one card of the plurality of cards is identified. Based on the events, a second card configuration is determined. Another content item selection event that involves the content delivery campaign is initiated and the content delivery campaign is selected. The content item is transmitted to a second computing device, where the multiple cards have the second card configuration.

Abstract: The disclosed technology teaches reducing threat detection processing, including recognizing that a file is an edited version of a previously processed file and retrieving, from an archive, metadata values, hashes for property groups and an entropy measure of the previously processed file. Also included is parsing the file into metadata values and property groups and calculating hashes of the property groups and entropy measure for the file. The method further includes applying similarity measures to compare the metadata values, the entropy measures, and the hashes on the property groups, for the edited version and the previously processed file. When any similarity measure or combination of similarity measures reaches a trigger, the technology teaches processing the file by using a threat detection module to detect malware. Property groups include core properties, application properties, document content and programming scripts for the edited version of the file and the previously processed file.

Abstract: Example embodiments described herein are directed towards dynamic state estimation of an operating state of a generator in a power system. Such estimation is performed for an individual generator in real time with improved accuracy and without the use of Global Position System (GPS) synchronization.

Abstract: The disclosed technology teaches reducing threat detection processing, including recognizing that a file is an edited version of a previously processed file and retrieving, from an archive, metadata values, hashes for property groups and an entropy measure of the previously processed file. Also included is parsing the file into metadata values and property groups and calculating hashes of the property groups and entropy measure for the file. The method further includes applying similarity measures to compare the metadata values, the entropy measures, and the hashes on the property groups, for the edited version and the previously processed file. When any similarity measure or combination of similarity measures reaches a trigger, the technology teaches processing the file by using a threat detection module to detect malware. Property groups include core properties, application properties, document content and programming scripts for the edited version of the file and the previously processed file.