Abstract: An identity document detector comprising a two-dimensional convolutional neural network is trained to detect categories of identity documents based on inputting unprocessed image data from documents. The documents comprise documents monitored by a data loss prevention (DLP) system across an organization, and each category of identity documents has associated risk levels for DLP. The DLP system performs corrective action to prevent data leakage based on detection of identity documents by the trained identity document detector and risk associated with the detected identity document categories as well as document context within the organization.

Abstract: A communication network comprising a plurality of CBSDs (Citizen Broadband Radio System Devices) and a Frequency Allocation System is disclosed. In some embodiments, the Frequency Allocation System includes a SON (self-organizing network). A SON, in accordance with the disclosed method and apparatus, organizes a network to have an outer and inner ring configuration of CBSDs that enhances strategic planning for the spectrum being allocated for use by the network. The use of this configuration may significantly minimize co-channel interference between several frequency bands, such as DoD (Department of Defense) frequency bands 3.1 to 3.4 GHz; CBRS frequency bands 3.55 to 3.7 GHz; and WiFi frequency bands 5 GHz and 6 GHz.

Abstract: A differencing technique enables efficient retrieval of data from one of a substantial number of point-in-time images (e.g., snapshots) maintained over substantially long periods of time in a long-term storage service (LTSS) of an archival storage system. The LTSS efficiently retrieves the data by computing differences or deltas between any two arbitrary snapshots in accordance with a differencing procedure. According to the technique, the differencing procedure operates on one or more index tree structures configured to translate a logical offset range of snapshot data in a snapshot address space (e.g., of a file system) to a data object address space (e.g., of an object store hosting the snapshot data).

Abstract: Methods, systems, apparatuses, and computer program products are provided that enable immediate access to data associated with a source blob. The foregoing is achieved by creating a copy-on-read blob based on the source blob and a link blob based on the copy-on-read blob for each virtual machine to be deployed. When creating the copy-on-read blob, the source blob is not copied. Instead, data from the source blob is copied to the copy-on-read blob upon request by a virtual machine. Subsequent requests for that data are retrieved from the copy-on-read blob. Each link blob stores write data provided by its corresponding virtual machine. When a virtual machine requests data, its corresponding link blob is checked for the data. If the data is not stored therein, the copy-on-read blob is checked for the data. If the copy-on-read blob does not store the data, the data is retrieved from the source blob.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for realizing a multimodal image classifier. In an aspect, a method includes, for each image of a plurality of images: processing the image by a textual generator model to obtain a set of phrases that are descriptive of the content of the image, wherein each phrase is one or more terms, processing the set of phrases by a textual embedding model to obtain an embedding of predicted text for the image, and processing the image using an image embedding model to obtain an embedding of image pixels of the image. Then a multimodal image classifier is trained on the embeddings of predicted text for the images and the embeddings of image pixels for the images to produce, as output, labels of an output taxonomy to classify an image based on the image as input.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for realizing a multimodal image classifier. In an aspect, a method includes, for each image of a plurality of images: processing the image by a textual generator model to obtain a set of phrases that are descriptive of the content of the image, wherein each phrase is one or more terms, processing the set of phrases by a textual embedding model to obtain an embedding of predicted text for the image, and processing the image using an image embedding model to obtain an embedding of image pixels of the image. Then a multimodal image classifier is trained on the embeddings of predicted text for the images and the embeddings of image pixels for the images to produce, as output, labels of an output taxonomy to classify an image based on the image as input.

Abstract: Methods, systems, apparatuses, and computer program products are provided that enable immediate access to data associated with a source blob. The foregoing is achieved by creating a copy-on-read blob based on the source blob and a link blob based on the copy-on-read blob for each virtual machine to be deployed. When creating the copy-on-read blob, the source blob is not copied. Instead, data from the source blob is copied to the copy-on-read blob upon request by a virtual machine. Subsequent requests for that data are retrieved from the copy-on-read blob. Each link blob stores write data provided by its corresponding virtual machine. When a virtual machine requests data, its corresponding link blob is checked for the data. If the data is not stored therein, the copy-on-read blob is checked for the data. If the copy-on-read blob does not store the data, the data is retrieved from the source blob.

Abstract: Methods, systems, apparatuses, and computer program products are provided that enable immediate access to data associated with a source blob. The foregoing is achieved by creating a copy-on-read blob based on the source blob and a link blob based on the copy-on-read blob for each virtual machine to be deployed. When creating the copy-on-read blob, the source blob is not copied. Instead, data from the source blob is copied to the copy-on-read blob upon request by a virtual machine. Subsequent requests for that data are retrieved from the copy-on-read blob. Each link blob stores write data provided by its corresponding virtual machine. When a virtual machine requests data, its corresponding link blob is checked for the data. If the data is not stored therein, the copy-on-read blob is checked for the data. If the copy-on-read blob does not store the data, the data is retrieved from the source blob.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for realizing a multimodal image classifier. In an aspect, a method includes, for each image of a plurality of images: processing the image by a textual generator model to obtain a set of phrases that are descriptive of the content of the image, wherein each phrase is one or more terms, processing the set of phrases by a textual embedding model to obtain an embedding of predicted text for the image, and processing the image using an image embedding model to obtain an embedding of image pixels of the image. Then a multimodal image classifier is trained on the embeddings of predicted text for the images and the embeddings of image pixels for the images to produce, as output, labels of an output taxonomy to classify an image based on the image as input.

Abstract: A method and system for managing connections with a distributed control plane is provided. The method includes generating, by a router, a controller identifier (ID) list comprising a plurality of controller group IDs of a plurality of controller groups, wherein one controller group ID uniquely identifies one controller group. The method also includes identifying a first controller group, by the router from the list, with which a connection is to be established. Further, the method includes establishing, by the router, the connection with a controller of the first controller group if at least one of following conditions is met I) the router has not exhausted maximum number of connections, 2) the router has previously had a connection with the controller of the first controller group, and 3) the router has an existing connection with a controller of a second controller group not present in the list.

Abstract: Methods, systems, apparatuses, and computer program products are provided that enable immediate access to data associated with a source blob. The foregoing is achieved by creating a copy-on-read blob based on the source blob and a link blob based on the copy-on-read blob for each virtual machine to be deployed. When creating the copy-on-read blob, the source blob is not copied. Instead, data from the source blob is copied to the copy-on-read blob upon request by a virtual machine. Subsequent requests for that data are retrieved from the copy-on-read blob. Each link blob stores write data provided by its corresponding virtual machine. When a virtual machine requests data, its corresponding link blob is checked for the data. If the data is not stored therein, the copy-on-read blob is checked for the data. If the copy-on-read blob does not store the data, the data is retrieved from the source blob.

Abstract: Methods, systems, apparatuses, and computer program products are provided that enable immediate access to data associated with a source blob. The foregoing is achieved by creating a copy-on-read blob based on the source blob and a link blob based on the copy-on-read blob for each virtual machine to be deployed. When creating the copy-on-read blob, the source blob is not copied. Instead, data from the source blob is copied to the copy-on-read blob upon request by a virtual machine. Subsequent requests for that data are retrieved from the copy-on-read blob. Each link blob stores write data provided by its corresponding virtual machine. When a virtual machine requests data, its corresponding link blob is checked for the data. If the data is not stored therein, the copy-on-read blob is checked for the data. If the copy-on-read blob does not store the data, the data is retrieved from the source blob.

Abstract: A method and system for managing connections with a distributed control plane is provided. The method includes generating, by a router, a controller identifier (ID) list comprising a plurality of controller group IDs of a plurality of controller groups, wherein one controller group ID uniquely identifies one controller group. The method also includes identifying a first controller group, by the router from the list, with which a connection is to be established. Further, the method includes establishing, by the router, the connection with a controller of the first controller group if at least one of following conditions is met I) the router has not exhausted maximum number of connections, 2) the router has previously had a connection with the controller of the first controller group, and 3) the router has an existing connection with a controller of a second controller group not present in the list.

Abstract: A method and system for managing connections with a distributed control plane is provided. The method includes generating, by a router, a controller identifier (ID) list comprising a plurality of controller group IDs of a plurality of controller groups, wherein one controller group ID uniquely identifies one controller group. The method also includes identifying a first controller group, by the router from the list, with which a connection is to be established. Further, the method includes establishing, by the router, the connection with a controller of the first controller group if at least one of following conditions is met 1) the router has not exhausted maximum number of connections, 2) the router has previously had a connection with the controller of the first controller group, and 3) the router has an existing connection with a controller of a second controller group not present in the list.

Abstract: Methods, systems, apparatuses, and computer program products are provided that enable immediate access to data associated with a source blob. The foregoing is achieved by creating a copy-on-read blob based on the source blob and a link blob based on the copy-on-read blob for each virtual machine to be deployed. When creating the copy-on-read blob, the source blob is not copied. Instead, data from the source blob is copied to the copy-on-read blob upon request by a virtual machine. Subsequent requests for that data are retrieved from the copy-on-read blob. Each link blob stores write data provided by its corresponding virtual machine. When a virtual machine requests data, its corresponding link blob is checked for the data. If the data is not stored therein, the copy-on-read blob is checked for the data. If the copy-on-read blob does not store the data, the data is retrieved from the source blob.

Abstract: A method and system for managing connections with a distributed control plane is provided. The method includes generating, by a router, a controller identifier (ID) list comprising a plurality of controller group IDs of a plurality of controller groups, wherein one controller group ID uniquely identifies one controller group. The method also includes identifying a first controller group, by the router from the list, with which a connection is to be established. Further, the method includes establishing, by the router, the connection with a controller of the first controller group if at least one of following conditions is met 1) the router has not exhausted maximum number of connections, 2) the router has previously had a connection with the controller of the first controller group, and 3) the router has an existing connection with a controller of a second controller group not present in the list.

Abstract: An approach to generating device-specific configurations is described. In one approach, a method of generating a device-specific configuration for a target device is described. The method involves receiving a configuration parameter, and receiving command syntax information. A state description is generated from the configuration parameter, with reference to a configuration library. Device information is retrieved from the target device, and the device-specific configuration is generated with reference to the command syntax information, the device information, the state description, and a command library.

Abstract: Technology for network security is disclosed. In one embodiment, a method of managing network security includes receiving sampled packets. The sampled packets represent packets being sampled from network packet traffic in at least one location in a network. The sampled packets are converted into an appropriate format for analysis to form converted packets. Moreover, the converted packets are sent to a first group including at least one security device for analysis. If an event message is generated by the at least one security device as a result of analysis of the converted packets, the event message is received from the at least one security device. Network security is evaluated based on the event message and security policies and is adjusted based on that evaluation. The method may be implemented with a network manager.

Abstract: Technology for network security is disclosed. In one embodiment, a method of managing network security includes receiving sampled packets. The sampled packets represent packets being sampled from network packet traffic in at least one location in a network. The sampled packets are converted into an appropriate format for analysis to form converted packets. Moreover, the converted packets are sent to a first group including at least one security device for analysis. If an event message is generated by the at least one security device as a result of analysis of the converted packets, the event message is received from the at least one security device. Network security is evaluated based on the event message and security policies and is adjusted based on that evaluation. The method may be implemented with a network manager.

Abstract: Technology for network security is disclosed. In one embodiment, a method of managing network security includes receiving sampled packets. The sampled packets represent packets being sampled from network packet traffic in at least one location in a network. The sampled packets are converted into an appropriate format for analysis to form converted packets. Moreover, the converted packets are sent to a first group including at least one security device for analysis. If an event message is generated by the at least one security device as a result of analysis of the converted packets, the event message is received from the at least one security device. Network security is evaluated based on the event message and security policies and is adjusted based on that evaluation. The method may be implemented with a network manager.