Title: StreetGAN: towards road network synthesis with generative adversarial networks
Authors: Hartmann, Stefan
Weinmann, Michael
Wessel, Raoul
Klein, Reinhard
Citation: WSCG 2017: full papers proceedings: 25th International Conference in Central Europe on Computer Graphics, Visualization and Computer Visionin co-operation with EUROGRAPHICS Association, p. 133-142.
Issue Date: 2017
Publisher: Václav Skala - UNION Agency
Document type: konferenční příspěvek
URI: wscg.zcu.cz/WSCG2017/!!_CSRN-2701.pdf
ISBN: 978-80-86943-44-2
ISSN: 2464–4617 (print)
2464–4625 (CD-ROM)
Keywords: hluboké učení;generativní modelování;generativní nepřátelské sítě;generování silniční sítě
Keywords in different language: deep learning;generative modeling;generative adversarial networks;road network generation
Abstract: We propose a novel example-based approach for road network synthesis relying on Generative Adversarial Networks (GANs), a recently introduced deep learning technique. In a pre-processing step, we first convert a given representation of a road network patch into a binary image where pixel intensities encode the presence or absence of streets. We then train a GAN that is able to automatically synthesize a multitude of arbitrary sized street networks that faithfully reproduce the style of the original patch. In a post-processing step, we extract a graph-based representation from the generated images. In contrast to other methods, our approach does neither require domainspecific expert knowledge, nor is it restricted to a limited number of street network templates. We demonstrate the general feasibility of our approach by synthesizing street networks of largely varying style and evaluate the results in terms of visual similarity as well as statistical similarity based on road network similarity measures.
Rights: © Václav Skala - UNION Agency
Appears in Collections:WSCG 2017: Full Papers Proceedings

Files in This Item:
File Description SizeFormat 
Hartmann.pdfPlný text6,1 MBAdobe PDFView/Open

Please use this identifier to cite or link to this item: http://hdl.handle.net/11025/29554

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.