Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for generating trained neural network with increased robustness against adversarial attacks by utilizing a dynamic dropout routine and/or a cyclic learning rate routine. For example, the disclosed systems can determine a dynamic dropout probability distribution associated with neurons of a neural network. The disclosed systems can further drop neurons from a neural network based on the dynamic dropout probability distribution to help neurons learn distinguishable features. In addition, the disclosed systems can utilize a cyclic learning rate routine to force copy weights of a copy neural network away from weights of an original neural network without decreasing prediction accuracy to ensure that the decision boundaries learned are different.

Abstract: An electronic apparatus is provided, and includes a communicator and a processor. The communicator communicates with external devices, including a first external device and a second external device different from the first external device, for controlling operation of the electronic apparatus. The processor, based on a media signal corresponding to a content being received, processes the media signal for playing the content, based on a first control signal being received from the first external device while the media signal is processed, performs an operation corresponding to the received first control signal, based on authentication information being received from the second external device while the media signal is processed, transmits information corresponding to the content to the second external device, and based on a second control signal being received from the first external device or the second external device, performs an operation corresponding to the received second control signal.

Abstract: Methods, systems, and media for presenting interactive elements within video content are provided.

Abstract: The present invention provides a Multipath blockchain capability to verify blockchain data integrity based on transactions linked to entities that may be device-based or otherwise. Based on an entity's needs, in one embodiment, the Multipath blockchain capability enables the entity reading a block linked to the entity to directly point to only that entity's transactions. The entity's pointer to each path type may be associated with each data record. In an embodiment, each entity's transaction is a separate unit; not bundled with other entity's transactions. Each of these transactions or blocks may be linked to that entity's previous block.

Abstract: A method for detection of tampering in an executable code including one or more code blocks. The method includes monitoring execution of the executable code with a call stack data structure associated therewith, the execution involving accessing one or more address spaces; receiving information about the one or more address spaces, as accessed; comparing the received information about one or more accessed address spaces with information about one or more allowed address spaces defined in the call stack data structure of the executable code; raising a flag upon detection that the one or more accessed address spaces are different from the one or more allowed address spaces, based on the comparison; and executing an action based on the raised flag.

Abstract: A computer network includes a camera node, a network access node, a verification node, and a display node. Video content recorded by a camera at the camera node is transmitted to the display node and to the verification node for verification. The video content is verified at the display node and at the verification node. Recording metadata of the video content is stored in a distributed ledger and retrieved by the display node to verify the video content. The verification node receives, from the network access node, verification data for verifying the video content.

Abstract: On-demand systems and methods are provided to manage locally-stored on-demand content. A user's equipment controls the deletion of and/or access to on-demand content from a local storage device based on constraining criteria that may include rental conditions, dynamic factors, and keys.

Abstract: A method, a computing device, and a non-transitory machine-readable medium for detecting malware attacks. In one example, an agent implemented in an operating system detects an overwrite in which an original data component is overwritten with a new data component. The agent computes a plurality of features associated with the overwrite, the plurality of features including an original entropy corresponding to the original data component, a new entropy corresponding to the new data component, an overwrite fraction, and a set of divergence features. The agent determines whether the new data component is encrypted using the plurality of features.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for implementing a privacy manager are disclosed. In one aspect, a method includes the actions of receiving, from a client device, an indication of a first setting of the client device. The actions further include receiving, from the client device, an indication of a type of data that the client device is prepared to transmit. The actions further include, based on the first setting and the type of data, generating an instruction to adjust the first setting or a second setting of the client device. The actions further include, before the client device transmits the data, providing, for output to the client device, the instruction to adjust the first setting or the second setting of the client device.

Abstract: In an approach for computing trade-offs between privacy and accuracy of data analysis on building a learning model, a processor receives a dataset for training a model. The dataset includes one or more pre-identified sensitive data fields. The processor determines a weight of each sensitive data field for the model. The processor evaluates resource cost of applying a privacy preservation technique to the one or more pre-identified sensitive data fields. The processor identifies correlation among the sensitive data fields. The processor presents a comparison of options for training the model, in terms of tradeoffs of accuracy for training the model and the resource cost of the privacy preservation technique.

Abstract: A system and method for automatically updating biometric data for a user by preferably retaining an ongoing digital collection of images/biometric captures/IOT (“Internet of Things”) captures taken over the life of the user. Changes in appearance of the user due to age, health, facial hair, hair color, hair length and many other evolutionary changes are automatically captured and contribute to the learning set. Accordingly, the system itself gets smarter and faster with every subsequent authentication. In one non-limiting embodiment, a target of the most recent 500 captures are retained, and tuning may suggest a greater number of captures should or could be used. Preferably, with each new capture the oldest capture in the ongoing digital collection is automatically deleted by the system.

Abstract: Mechanisms are provided for detecting abnormal system call sequences in a monitored computing environment. The mechanisms receive, from a computing system resource of the monitored computing environment, a system call of an observed system call sequence for evaluation. A trained recurrent neural network (RNN), trained to predict system call sequences, processes the system call to generate a prediction of a subsequent system call in a predicted system call sequence. Abnormal call sequence logic compares the subsequent system call in the predicted system call sequence to an observed system call in the observed system call sequence and identifies a difference between the predicted system call sequence and the observed system call sequence based on results of the comparing. The abnormal call sequence logic generates an alert notification in response to identifying the difference.

Abstract: Methods and systems for filtering content in reconstructions of native data in assets. Specifically, the methods and systems extract telemetry data from native game data. The methods and systems then identify Personally Identifiable Information (“PII”) in the telemetry data and determine a masking operation to perform on the PII. The methods and systems then generate for presentation, in a user interface, reconstructed asset data based on the telemetry data, wherein the PII is modified in the reconstructed asset data based on the masking operation.

Abstract: Systems, computer program products, and methods are described herein for secure access control using dynamic resource replication. The present invention is configured to electronically receive, from a computing device of a user, a request to generate a resource access path to access a resource; determine one or more resource requirements associated with the resource; determine an authentication level associated with the user; initiate a resource replication engine on the one or more resource requirements and the authentication level associated with the user; determine, using the resource replication engine, that the one or more requirements and the authentication level associated with the user meets one or more conditions for a replication process; and generate, using the resource replication engine, the resource access path and a plurality of replicate resource access paths.

Abstract: A method of block chain based data protection may comprise receiving a user block chain instruction to record a address-identified memory location at which a dataset field value containing sensitive personal information is stored pursuant to a customized data integration process modeled via a graphical user interface, creating a block chain associated with the dataset field value, receiving an identification of the address-identified memory location from a customized data integration process remote execution location, and creating a first block storing the identification of the address-identified memory location within the block chain.

Abstract: Techniques are shown for key management using a traceable key block-chain ledger involving creating a cryptographic key at a key source, generating a key block on a block-chain ledger corresponding to the cryptographic key, and securely modifying the key block to include metadata describing the key source. The techniques also involve performing a first key transaction with the cryptographic key, generating a first transaction block on the block-chain ledger corresponding to the first key transaction with the cryptographic key, linking the first transaction block to the key block and securely modifying the first transaction block to include metadata describing the first key transaction with the cryptographic key.

Abstract: A method may include receiving a registration request at a server of a consent management platform from a content-presentation device, and using an authentication certificate in the request to establish a secure communicative connection. The server may generate: a global ID (GID) from information received over the secure connection; a device-based device record for the device, and including the GID and a unique address indicator; and a cryptographically-signed token. The GID, device record, and token may be transmitted to the device. The unique address indicator may be associated with consent packages having features of a media distribution system that require user consent to associated agreements for activation on the device. The server may generate a server-based device record duplicating the device-based device record, and including the consent packages and indicators of consent agreement status initialized to undeclared. The server may store the server-based device record in a flat database.

Abstract: A method and system for planning and running video-on-demand (VOD) advertising includes a planner for planning advertising campaigns and an ad selector for running selected advertisements from selected ones of the campaigns. The campaigns are associated with household demographic classifications, video content category categories, household geographic locations, and/or video viewing times. In response to a viewer requesting a video, the selector selects the campaigns corresponding to the demographic classification associated with the viewer, the content category of the requested video, the geographic location associated with the viewer, and/or the viewing time at which the requested video is viewed. The demographic classification is determined by using an identifier of the viewer's set-top-box. The selector may select the campaigns corresponding to the content categories of the requested video. The content categories are determined by using an identifier associated with the requested video.

Abstract: An embodiment of the present invention is directed to delivering an entitlements model that scales to both mid-frequency and low-latency use cases. The innovative solution may be distributed in nature and able to operate in low priority threads alongside the main logic of the software. An embodiment of the present invention may be implemented as a software module with APIs for ease of adoption.

Abstract: A system and method for executing and recording transactions for a key in a blockchain. A method includes creating a smart contract on receiving a request to order a key on lease for a first period of time. The method further includes determining an availability of the key and determining an availability of digital tokens in an account of the user. The method also includes executing the smart contract for the first period of time by executing a first transaction using the digital tokens and recording the executed first transaction and the executed smart contract in a block of a block chain to provide the key on lease.