Abstract: Graph Convolution Networks (GCNs) have achieved remarkable success in representation of structured graph data. As we know that traditional GCNs are generally defined on the fixed first-order ...
Abstract: Deep learning has emerged as the predominant approach for ship detection in synthetic aperture radar (SAR) imagery. Nevertheless, persistent challenges such as densely clustered vessels, ...
We introduce a novel generative model: Discrete Distribution Networks (DDN). DDN employs a simple yet unique approach to model the target distribution, distinctly different from mainstream generative ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results