Adaptive Sparsified Graph Learning Framework for Vessel Behavior Anomalies
This addresses the challenge of modeling spatiotemporal interactions in dynamic environments like maritime traffic, which is incremental over existing graph neural network methods.
The paper tackles the problem of detecting vessel behavior anomalies in dynamic maritime environments by introducing a graph representation where timestamps are modeled as nodes to capture temporal dependencies, achieving robust anomaly detection through a forecasting layer and Variational Graph Autoencoder.
Graph neural networks have emerged as a powerful tool for learning spatiotemporal interactions. However, conventional approaches often rely on predefined graphs, which may obscure the precise relationships being modeled. Additionally, existing methods typically define nodes based on fixed spatial locations, a strategy that is ill-suited for dynamic environments like maritime environments. Our method introduces an innovative graph representation where timestamps are modeled as distinct nodes, allowing temporal dependencies to be explicitly captured through graph edges. This setup is extended to construct a multi-ship graph that effectively captures spatial interactions while preserving graph sparsity. The graph is processed using Graph Convolutional Network layers to capture spatiotemporal patterns, with a forecasting layer for feature prediction and a Variational Graph Autoencoder for reconstruction, enabling robust anomaly detection.