Method and device for analyzing battlefield threat based on dynamic graph neural network
AssigneeAgency for Defense Development
InventorsSangheun SHIM, Dong Kuk RYU, Kiwoong PARK
Life of this patent
Estimating the term…
Abstract
Proposed is a method and device for analyzing battlefield threat based on a dynamic graph neural network. The method may include classifying a behavior at a specific time point for a plurality of objects in a battlefield based on battlefield information received from the outside. The method may also include predicting a threat occurrence probability between the plurality of objects according to the classified behavior. The method may further include generating and outputting a battlefield threat analysis result including the classified behavior and the threat occurrence probability.
Claims15
A method for analyzing a battlefield threat, the method comprising:
- receiving battlefield information including object information and edge information for each of a plurality of objects in a battlefield and generating a plurality of time point-specific graph data for the plurality of objects;
- calculating a plurality of time point-specific latent vectors for each object from the plurality of time point-specific graph data of each object using a previously learned neural network model;
- and
- classifying a behavior of each of the plurality of objects from the plurality of time point-specific latent vectors using the neural network model, and predicting a threat occurrence probability between the plurality of objects at a future time point based on the classified behavior.
A device for analyzing a battlefield threat, the device comprising:
- a memory configured to store a battlefield threat analysis program;
- and
- one or more processors configured to execute the battlefield threat analysis program to:
- receive battlefield information including object information and edge information for each of a plurality of objects in a battlefield and generate a plurality of time point-specific graph data for the plurality of objects,
- calculate a plurality of time point-specific latent vectors for each object from the plurality of time point-specific graph data of each object using a previously learned neural network model, and
- classify a behavior of each of the plurality of objects from the plurality of time point-specific latent vectors using the neural network model, and predict a threat occurrence probability between the plurality of objects at a future time point based on the classified behavior.
A non-transitory computer-readable recording medium storing instructions, when executed by one or more processors, that cause the one or more processors to perform a method for analyzing a battlefield threat including:
- receiving battlefield information including object information and edge information for each of a plurality of objects in a battlefield and generating a plurality of time point-specific graph data for the plurality of objects;
- calculating a plurality of time point-specific latent vectors for each object from the plurality of time point-specific graph data of each object using a previously learned neural network model;
- and
- classifying a behavior of each of the plurality of objects from the plurality of time point-specific latent vectors using the neural network model, and predicting a threat occurrence probability between the plurality of objects at a future time point based on the classified behavior.
Dependencies are inferred from the claim text.