LayoutGKN: Graph Similarity Learning of Floor Plans
This addresses efficiency issues for applications like search and clustering in architectural or design domains, but it is incremental as it builds on existing graph matching methods.
The paper tackled the problem of slow inference time in graph matching networks for comparing floor plan graphs by introducing LayoutGKN, which postpones cross-graph interactions to the end, resulting in comparable or better similarity computation with significantly increased speed.
Floor plans depict building layouts and are often represented as graphs to capture the underlying spatial relationships. Comparison of these graphs is critical for applications like search, clustering, and data visualization. The most successful methods to compare graphs \ie, graph matching networks, rely on costly intermediate cross-graph node-level interactions, therefore being slow in inference time. We introduce \textbf{LayoutGKN}, a more efficient approach that postpones the cross-graph node-level interactions to the end of the joint embedding architecture. We do so by using a differentiable graph kernel as a distance function on the final learned node-level embeddings. We show that LayoutGKN computes similarity comparably or better than graph matching networks while significantly increasing the speed. \href{https://github.com/caspervanengelenburg/LayoutGKN}{Code and data} are open.