Abstract: Devices, systems, and methods for machine learning (ML) automatic target recognition (ATR) decision explanation are provided. A method can include receiving an object specification matrix from an object model database that indicates, for each of a plurality of physical portions of an object, whether each of a plurality of features are present or absent in a physical portion of the physical portions of the object and a proportional physical displacement between the features in the object, receiving feature data indicating for an image of a portion of the object, a likelihood whether each of features are present in the image, determining based on the object specification matrix and the feature data, a probability and corresponding uncertainty that the image corresponds to the object, and providing the probability and corresponding uncertainty of the object to help classify the object.

Abstract: A multi-domain operational system is provided. The system includes a data collection engine configured and operable to collect operational data from one or more collection resources within time constraints imposed by operational requirements. The system further includes a business rules engine configured and operable to execute a plurality of rules each of which comprise a plurality of conditions grouped in a sequence by a set of logical connectors. The plurality of rules are associated with a mission task. The system also includes a security engine configured to facilitate information sharing across multiple security domains based on one or more security protocols. The system further includes a mission manager software component configured and operable to manage operation of at least the data collection engine, the business rules engine and the security engine using a common information layer.

Abstract: A method includes receiving a directive from a user to find an object in a geographical area, wherein the object is identified with an input label selected from a set of labels, obtaining sensor data in response to the directive for a real world physical object in the geographical area using one or more sensors, processing the sensor data with a plurality of automatic target recognition (ATR) algorithms to assign a respective ATR label from the set of labels and a respective confidence level to the real world physical object, and receiving modeled relationships within the set of labels using a probabilistic model based on a priori knowledge encoded in a set of model parameters. The method includes inferring an updated confidence level that the real world physical object actually corresponds to the input label based on the ATR labels and confidences and based on the probabilistic model.

Abstract: A method is disclosed of providing a graphical user interface (GUI) for simulating target detection and recognition. The method includes providing a user entry area for receiving user entered ground truth attributes defining one or more respective simulated physical objects, and providing a user entry area for receiving user entered collection attributes defining one or more collections, each collection being a simulation of automated target recognition applied to simulated sensor output associated with simulated sensing of the one or more simulated physical objects. The method further includes simulating detection of the one or more simulated physical objects and recognition based on the ground truth attributes and the collection attributes, generating detection data about at least one detected object based on the simulated target detection, and displaying the detection data.

Abstract: A method includes receiving a query from a user requesting a prediction on future location of a target at a time t>0, receiving a trafficability map wherein the target is located in the trafficability map at an initial time t=0, receiving information about the target including uncertainty on speed of the target, modeling motion of the target using information from the trafficability map including terrain and target mobility on different terrain types to locations on the trafficability map, and answering the query from the user based on output from modeling motion of the target.

Abstract: A simulator to simulate target detection and recognition is provided. The simulator includes a memory configured to store instructions and a processor disposed in communication with said memory. The processor upon execution of the instructions is configured to receive ground truth attributes defining one or more respective simulated physical objects and receive collection attributes defining one or more collections. Each collection is a simulation of automated target recognition applied to simulated sensor output associated with simulated sensing of the one or more simulated physical objects. The processor upon execution of the instructions is further configured to simulate detection of the one or more simulated physical objects and recognition based on the ground truth attributes and the collection attributes and output detection data about at least one detected object based on the simulation of target detection.

Abstract: A method to simulate target detection and recognition is provided. The method includes receiving ground truth attributes defining one or more respective simulated physical objects and receiving collection attributes defining one or more collections. Each collection is a simulation of automated target recognition applied to simulated sensor output associated with simulated sensing of the one or more simulated physical objects. The method further includes simulating detection of the one or more simulated physical objects and recognition based on the ground truth attributes and the collection attributes, generating detection data about at least one detected object based on the simulated target detection, and outputting the detection data.

Abstract: A method includes receiving a directive from a user to find an object in a geographical area, wherein the object is identified with an input label selected from a set of labels, obtaining sensor data in response to the directive for a real world physical object in the geographical area using one or more sensors, processing the sensor data with a plurality of automatic target recognition (ATR) algorithms to assign a respective ATR label from the set of labels and a respective confidence level to the real world physical object, and receiving modeled relationships within the set of labels using a probabilistic model based on a priori knowledge encoded in a set of model parameters. The method includes inferring an updated confidence level that the real world physical object actually corresponds to the input label based on the ATR labels and confidences and based on the probabilistic model.

Abstract: A method includes receiving a query from a user requesting a prediction on future location of a target at a time t>0, receiving a trafficability map wherein the target is located in the trafficability map at an initial time t=0, receiving information about the target including uncertainty on speed of the target, modeling motion of the target using information from the trafficability map including terrain and target mobility on different terrain types to locations on the trafficability map, and answering the query from the user based on output from modeling motion of the target.

Abstract: A method for information fusion in multi-domain operational environment is provided. The method includes receiving a first plurality of detection tags containing domain agnostic information related to one or more targets of interest. Target detection statistics contained in the first plurality of detection tags are combined to provide a fused estimate of a probability of a correct identification of a target of the one or more targets of interest. Target geolocation statistics contained in the first plurality of detection tags are combined to provide a fused estimate of a location of the one or more targets of interest. One or more fused detection tag containing at least the combined target detection statistics and the combined target geolocation statistics are generated.

Abstract: A system and computer-implemented method for automatically recognizing a new class in a classification system. The method includes accessing components of a trained convolutional neural network (CNN) that has been trained with available classes. The components are provided in a kernel space and include at least one of a plurality of kernels and a plurality of neurons of one or more layers of the CNN. Furthermore, the components are assigned to a class in accordance with the training. The method further includes applying a covariance matrix to map the components in the kernel space to eigenspace; determining, for each of the available classes, an eigen-distance between a sample and the components mapped to eigenspace; based on the eigen-distance, determining whether the sample is an outlier that does not belong to one of the classes; and creating a new class that includes the sample if determined that the sample is an outlier.

Abstract: A method for automated operational services across multiple data domains includes receiving requirement information related to one or more targets of interest. Information related to the one or more targets of interest is collected from one or more data sources across multiple data domains based on the received requirement information. A determination is made if the collected information meets the received requirements related to the one or more targets of interest. A classification tag containing domain agnostic information related to the one or more targets of interest is generated, responsive to determining that the collected information meets the received requirements.

Abstract: A method of predicting warps for image registration includes receiving warp data for a first band pair in an N-band system. The method includes fitting a warp model to the warp data to produce an offset profile for the first band pair, predicting a respective offset profile for at least one other band pair in the N-band system, and using the predicted respective offset profile to generate a warp for the at least one other band pair. The method also includes registering images of the at least one other band pair using the warp for the at least one other band pair. In another aspect, a method of predicting warps for image registration includes receiving warp data for m band pairs in an N-band system, wherein m?(N?1) and m>2, to register images.

Abstract: A system and computer-implemented method for automatically recognizing a new class in a classification system. The method includes accessing components of a trained convolutional neural network (CNN) that has been trained with available classes. The components are provided in a kernel space and include at least one of a plurality of kernels and a plurality of neurons of one or more layers of the CNN. Furthermore, the components are assigned to a class in accordance with the training. The method further includes applying a covariance matrix to map the components in the kernel space to eigenspace; determining, for each of the available classes, an eigen-distance between a sample and the components mapped to eigenspace; based on the eigen-distance, determining whether the sample is an outlier that does not belong to one of the classes; and creating a new class that includes the sample if determined that the sample is an outlier.

Abstract: A method to simulate target detection and recognition is provided. The method includes receiving ground truth attributes defining one or more respective simulated physical objects and receiving collection attributes defining one or more collections. Each collection is a simulation of automated target recognition applied to simulated sensor output associated with simulated sensing of the one or more simulated physical objects. The method further includes simulating detection of the one or more simulated physical objects and recognition based on the ground truth attributes and the collection attributes, generating detection data about at least one detected object based on the simulated target detection, and outputting the detection data.

Abstract: A simulator to simulate target detection and recognition is provided. The simulator includes a memory configured to store instructions and a processor disposed in communication with said memory. The processor upon execution of the instructions is configured to receive ground truth attributes defining one or more respective simulated physical objects and receive collection attributes defining one or more collections. Each collection is a simulation of automated target recognition applied to simulated sensor output associated with simulated sensing of the one or more simulated physical objects. The processor upon execution of the instructions is further configured to simulate detection of the one or more simulated physical objects and recognition based on the ground truth attributes and the collection attributes and output detection data about at least one detected object based on the simulation of target detection.

Abstract: A method is disclosed of providing a graphical user interface (GUI) for simulating target detection and recognition. The method includes providing a user entry area for receiving user entered ground truth attributes defining one or more respective simulated physical objects, and providing a user entry area for receiving user entered collection attributes defining one or more collections, each collection being a simulation of automated target recognition applied to simulated sensor output associated with simulated sensing of the one or more simulated physical objects. The method further includes simulating detection of the one or more simulated physical objects and recognition based on the ground truth attributes and the collection attributes, generating detection data about at least one detected object based on the simulated target detection, and displaying the detection data.

Abstract: A method for automated operational services across multiple data domains includes receiving requirement information related to one or more targets of interest. Information related to the one or more targets of interest is collected from one or more data sources across multiple data domains based on the received requirement information. A determination is made if the collected information meets the received requirements related to the one or more targets of interest. A classification tag containing domain agnostic information related to the one or more targets of interest is generated, responsive to determining that the collected information meets the received requirements.

Abstract: A multi-domain operational system is provided. The system includes a data collection engine configured and operable to collect operational data from one or more collection resources within time constraints imposed by operational requirements. The system further includes a business rules engine configured and operable to execute a plurality of rules each of which comprise a plurality of conditions grouped in a sequence by a set of logical connectors. The plurality of rules are associated with a mission task. The system also includes a security engine configured to facilitate information sharing across multiple security domains based on one or more security protocols. The system further includes a mission manager software component configured and operable to manage operation of at least the data collection engine, the business rules engine and the security engine using a common information layer.

Abstract: A method for information fusion in multi-domain operational environment is provided. The method includes receiving a first plurality of detection tags containing domain agnostic information related to one or more targets of interest. Target detection statistics contained in the first plurality of detection tags are combined to provide a fused estimate of a probability of a correct identification of a target of the one or more targets of interest. Target geolocation statistics contained in the first plurality of detection tags are combined to provide a fused estimate of a location of the one or more targets of interest. One or more fused detection tag containing at least the combined target detection statistics and the combined target geolocation statistics are generated.