English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Automatic building extraction from optical imagery remains a challenge dueto, for example, the complexity of building shapes. Semantic segmentation is anefficient approach for this task. The latest development in deep convolutionalneural networks (DCNNs) has made accurate pixel-level classification taskspossible. Yet one central issue remains: the precise delineation of boundaries.Deep architectures generally fail to produce fine-grained segmentation withaccurate boundaries due to their progressive down-sampling. Hence, we introducea generic framework to overcome the issue, integrating the graph convolutionalnetwork (GCN) and deep structured feature embedding (DSFE) into an end-to-endworkflow. Furthermore, instead of using a classic graph convolutional neuralnetwork, we propose a gated graph convolutional network, which enables therefinement of weak and coarse semantic predictions to generate sharp bordersand fine-grained pixel-level classification. Taking the semantic segmentationof building footprints as a practical example, we compared different featureembedding architectures and graph neural networks. Our proposed framework withthe new GCN architecture outperforms state-of-the-art approaches. Although ourmain task in this work is building footprint extraction, the proposed methodcan be generally applied to other binary or multi-label segmentation tasks.

Building Segmentation through a Gated Graph Convolutional Neural Network with Deep Structured Feature Embedding