Abstract: Embodiments relate to a digital image file alteration detection controller which implements a processor configuration to efficiently detect an alteration to a digital image file. The digital image file alteration detection controller can include a processor, and a memory associated with the processor, the memory including instructions stored thereon that when executed by the processor will cause the processor to: extract Photo Response Non-Uniformity (PRNU) data of a digital image file received from the memory; determine a local variability representing a variability in PRNU data for a locale of a digital image file; determine a global variability representing a variability in PRNU data for an entire digital image file; compare local variability to global variability; and generate an alteration detection indicator indicative of an alteration detected when the local variability to global variability comparison is less than a threshold value.

Abstract: An exemplary mobile apparatus for assessing an operating condition of an asset. The mobile device includes at least one sensor device for acquiring data related to one or more operational or environmental characteristics of an asset during operation. The mobile device further includes a processing device encoded with a neural network architecture having one or more models trained to identify one or more operating conditions of one or more assets according to asset type. The processing device can determine an operating condition of the asset by extracting features from the acquired data and comparing attributes of the extracted features to attributes of at least one known operating condition determined through training of the models. The mobile device can also include an output interface to output a result of the determination performed by the processing device, which can include communicating the result to a remote device.

Abstract: A vehicle can be configured to include a body having a body bottom conjoined with a body sidewall and a body top forming a body cavity. The body top includes a body top opening and the body sidewall includes a body sidewall opening. The vehicle can include a payload housing having a payload bottom conjoined with a payload housing sidewall and a payload housing top forming a payload housing cavity, wherein the payload housing cavity is configured to hold at least one operating module for the vehicle. The vehicle can include at least one arm. The vehicle can include at least one interlocking arrangement of the body top opening or body side wall configured to removably secure the payload housing and the at least one arm to the body. Each of the body, the payload housing, and the at least one arm can be structured with additive manufactured material.

Abstract: Embodiments can relate to a system for automated exploit generation that receives input data representative of a target action to establish a target having a potential target vulnerability. The system can build a simulated target environment that includes the established target. The system can conduct an analysis method including a static, a concrete, a dynamic, and/or a symbolic analysis. The system can create a chainable sequence including an information disclosure, a read, a write, and/or an execution exploit primitive. The system can generate an exploit chain that, when executed by the processor in response to the target action, can transform the target action to a target failure within the simulated target environment and thereby expose the target vulnerability. The system can execute the exploit chain within the simulated target environment to examine coverage of the exposed target vulnerability. The system can generate an output representative of the exposed target vulnerability.

Abstract: Provided are methods, systems, and non-transitory computer-readable media for generating a feature vector for malware, including storing, in memory of a computing device, program code for a trained neural network that produces embedded representations for antivirus scan data; executing, by a processor of the computing device, the program code for the trained neural network to perform the operations of: (a) receiving an antivirus scan report (AVSR) for a malware file; (b) normalizing each label in the AVSR by separating the label into a sequence of tokens including a set of token strings; (c) embedding a first token and plural second tokens to generate an input sequence for the malware file; (d) inputting the input sequence into a neural model for producing antivirus scan data; and (e) outputting the antivirus scan data produced by the neural model as one or more feature vectors.

Abstract: Provided are systems, methods, and computer program products including at least one processor programmed or configured to perturb at least one training dataset based on mutual information extracted from an ensemble machine learning model to provide at least one adversarial training dataset, execute at least two machine learning models of an ensemble machine learning model, train at least two machine learning models with the at least one training dataset by feeding an input or output of one of the at least two machine learning models to the other of the at least two machine learning models, train the ensemble machine learning model with the at least one adversarial training dataset, receive a runtime input from a client device, and provide the runtime input to the trained ensemble machine learning model to generate a signal output indicating that the runtime input includes an out-of-distribution sample.

Abstract: Systems for co-site interference mitigation are provided herein. A system according to an embodiment receives first spectrum usage data indicating parameters to be used by a first group of one or more radio elements to send or receive transmissions. The system receives second spectrum usage data indicating parameters to be used by a second group of one or more radio elements to send or receive transmissions. The system dynamically determines, based on the first and second spectrum usage data received in real time, that interference will occur between the first group and the second group. The system forwards, to the at least one of the radio elements, a signal indicating parameters for mitigating the interference between transmissions to be sent or received by the radio elements of the first or second group. The system sends or receives a transmission in accordance with the sent signal.

Abstract: Methods and devices for detection and geolocation of a target device in three-dimensional space are provided. A method may include capturing baseline information in an area of operation and receiving information associated with one or more detected signals from a target device within the area of operation. The one or more detected signals from the target device include information indicating a device identifier associated with the target device. The method may include filtering the baseline information from the one or more detected signals from the target device based on the indicated device identifier. The method may include calculating a three-dimensional location of the device based on a signal strength of the detected one or more signals and a plurality of three-dimensional locations where the one or more signals were detected. The method may include plotting the three-dimensional location of the target device on a map within a graphical user interface (GUI).

Abstract: Exemplary systems and methods are directed to embedding data into a machine learning model. A processing device executes program code for running a machine learning model, which has a plurality of parameter values. The processing device receives a message to be embedded into the machine learning model. The message is encrypted according to a set of keys of a cryptographic algorithm. The encrypted message is converted to a corresponding binary representation. The binary representation of the encrypted message is embedded into at least one of the one or more parameters of the machine learning model. The embedding operation modifies the at least one parameter value of the machine learning model.

Abstract: A method for generating trust metrics for sensor data is disclosed. The method can include receiving the sensor data from at least one sensor associated with a platform; categorizing the sensor data into one or more sensor data types; applying one or more threat detection algorithms to the sensor data based on the one or more sensor data types to detect one or more threats to the integrity of the sensor data; calculating a detection certainty for the sensor data from the at least one sensor, the detection certainty indicating a probability that the one or more threats are affecting the integrity of the sensor data; and generating a trust metric for the sensor data of the at least one sensor based on the detection certainty for the sensor data from the at least one sensor.

Abstract: Exemplary systems and methods are directed to training a machine learning model and for preventing leakage of training data by the machine learning model subsequent to training. A processor is configured to convert a sparse dataset into a matrix of plural data coordinates, generate a priority queue populated with the plural data coordinates, and iteratively select a data coordinate from the priority queue. Plural model values are calculated such that any zero value in the sparse dataset is avoided while maintaining a same result. A next feature is selected, and its weight is altered. Plural variables of the matrix are updated based on the altered weight value, and the priority queue is updated to adjust a priority of the data coordinates based on the update to the plural variables. The process is repeated for each next data coordinate until the model converges to a solution based on the model weights.

Abstract: A communication system is disclosed. The communication system includes a communication device comprising one of a receiver, a transmitter, or a transceiver; and an acoustic lensing subsystem, wherein the acoustic lensing subsystem is configured to: convert electrical signals into acoustic signals; focus the acoustic signals to generate focused acoustic signals; and output the focused acoustic signals.

Abstract: Embodiments relate to a detection system. The detection system includes a light detection and ranging (LIDAR) module configured to scan a swarm of objects to generate image data of a first object associated with a swarm. The detection system includes an image processing module configured to process the image data and control the at least one LIDAR module. The image processing module is configured to: detect presence of a first object; detect a feature of a first object for which the presence has been detected; characterize, using image processing, a feature of a first object; and initiate, based on the characterization of a feature, the LIDAR module to any one or combination of track a first object for which the presence has been detected or scan a swarm to generate image data of a second object associated with a swarm.

Abstract: Disclosed are methods and apparatuses for classifier evaluation. The evaluation involves constructing a ground truth refinement having a degree of error within specified bounds from a malware reference dataset as an approximate ground truth refinement. The evaluation further involves using the approximate ground truth refinement to determine at least one of: a lower bound on precision or an upper bound on recall and accuracy. The evaluation further involves evaluating a classifier by evaluating at least one of a classification method or clustering method by examining changes to the upper bound and/or the lower bound produced by the approximate ground truth refinement.

Abstract: A system and method for transferring an adversarial attack involving generating a surrogate model having an architecture and a dataset that mirrors at least one aspect of a target model of a target module, wherein the surrogate model includes a plurality of classes. The method involves generating a masked version of the surrogate model having fewer classes than the surrogate model by randomly selecting at least one class of the plurality of classes for removal. The method involves attacking the masked surrogate model to create a perturbed sample. The method involves generalizing the perturbed sample for use with the target module. The method involves transferring the perturbed sample to the target module to alter an operating parameter of the target model.

Abstract: Exemplary systems and methods are directed to generating customized imagery includes receiving input parameters that define operations for one of plural disparate image processing tools in generating the customized imagery and define attributes of the customized imagery to be generated. Program code for generating an API is executed and the API establishes communication with each image processing tool. The API generates parameterized calls which provide instructions for a specified one of the image processing tools to generate the customized imagery. The image processing tool which receives the instructions is identified from the input parameters. The parameterized calls are sent to the parameterized calls to the image processing tool and the customized imagery is generated. The customized imagery is returned to the API and is stored in a database as training data for an artificial intelligence model.

Abstract: Exemplary systems and methods are directed to risk-based observability of a platform. Data is received from plural devices from one or more computing environments on a network. The received data is in a raw data format according to the computing environment or platform from which it was received. The received data is converted from the raw format to a structured format. The converted data is enhanced by adding contextual information associated with a corresponding one of the plural devices. A risk analysis is performed on the enhanced data based on one or more risk detection rules applied to the network. One or more tags are applied to the enhanced data based on results of the risk analysis. Data analysis is performed on the enhanced data to identify devices from aggregate sources. The data is sent to one or more destinations on the network based on the applied tags.

Abstract: Thermal management techniques include: transporting a refrigerant fluid from a receiver to an inlet of a flash tank that has a vapor-side outlet and liquid-side outlet such that a liquid phase of the refrigerant fluid moves to a bottom of the flash tank and outputs from the liquid-side outlet; forming a solid-vapor state from the liquid phase by expanding the liquid phase with an expansion valve to a first pressure that is less than a triple point pressure to form a solid-vapor mixture of the refrigerant fluid; extracting heat from a heat load with an evaporator that receives the solid-vapor mixture of the refrigerant fluid and sublimates the solid state of the solid-vapor mixture of the refrigerant fluid directly into a vapor phase of the refrigerant fluid; and discharging, from an exhaust line, the vapor phase to an ambient environment without returning the vapor phase to the receiver.

Abstract: Exemplary systems and methods are directed to endpoint detection and response (EDR) in which a receiver receives streaming data from plural EDR platforms with vendor-specific data formats for the streaming data. An application programming interface converts the streaming data received from each EDR platform to a common data format. A detection engine analyzes the converted streaming data for attributes of malicious activity and generates an alert when malicious activity is detected. A graphical user interface filters and sorts the generated alerts based on at least one of a priority of addressing the malicious activity and a severity of harm caused by the malicious activity. The graphical user interface further generates an interactive display of the filtered and sorted alerts, where each alert includes an active or activatable link which when selected provides additional information obtained from one of the plural EDR platforms associated with the alert.

Abstract: A thermal management system includes an open-circuit refrigeration system including a cooling system configured to supply a cooling medium. The open-circuit refrigeration system includes a receiver having a receiver outlet, the receiver configurable to store a refrigerant fluid, the receiver configured to receive the cooling medium from the cooling system, an evaporator coupled to the receiver outlet, the evaporator configurable to receive liquid refrigerant fluid from the receiver outlet and to extract heat from a heat load when the heat load contacts or is proximate to the evaporator a control device configurable to control a temperature of the heat load and an exhaust line, with the receiver, the evaporator, and the exhaust line coupled to form an open-circuit refrigerant fluid flow path.