Das Paper "A Review and Efficient Implementation of Scene Graph Generation Metrics" von Julian Lorenz, Robin Sch?n, Katja Ludwig und Rainer Lienhart wurde beim Workshop on Scene Graphs and Graph Representation Learning auf der CVPR 2024 akzeptiert.

Die Autoren schaffen einen ?berblick über existierende Scene Graph Generation Metriken und pr?sentieren pr?zise Definitionen, die bisher noch nicht gegeben waren. Au?erdem stellen die Autoren ein Pythonpaket vor, dass eine effiziente und leicht zu verwendende Implementierung der eingeführten Metriken bietet. Um Scene Graph Generation Methoden in Zukunft besser vergleichen zu k?nnen wird ein Benchmarking-Service eingerichtet, über den neue Methoden leicht verglichen werden k?nnen.

Weitere Informationen sind unter https://lorjul.github.io/sgbench/ zu finden.

?

Scene graph generation has emerged as a prominent research field in computer vision, witnessing significant advancements in the recent years. However, despite these strides, precise and thorough definitions for the metrics used to evaluate scene graph generation models are lacking. In this paper, we address this gap in the literature by providing a review and precise definition of commonly used metrics in scene graph generation. Our comprehensive examination clarifies the underlying principles of these metrics and can serve as a reference or introduction to scene graph metrics.
Furthermore, to facilitate the usage of these metrics, we introduce a standalone Python package called SGBench that efficiently implements all defined metrics, ensuring their accessibility to the research community. Additionally, we present a scene graph benchmarking web service, that enables researchers to compare scene graph generation methods and increase visibility of new methods in a central place.

?