SCENE: Reasoning About Traffic Scenes Using Heterogeneous Graph Neural Networks

Understanding traffic scenes requires considering heterogeneous information about dynamic agents and the static infrastructure. In this work we propose SCENE, a methodology to encode diverse in graphs reason these using Graph Neural Network encoder task-specific decoders. The graphs, whose structures are defined by an ontology, consist of different nodes with type-specific node features relations edge features. order exploit all given use cascaded layers graph convolution. result is encoding scene. Task-specific decoders can be applied predict desired attributes Extensive evaluation on two binary classification tasks show main strength methodology: despite being generic, it even manages outperform baselines. further application our task various knowledge shows its transferability other domains.