Full metadata record
DC poleHodnotaJazyk
dc.contributor.authorFeng, Qi
dc.contributor.authorShum, Hubert P. H.
dc.contributor.authorShimamura, Ryo
dc.contributor.authorMorishima, Shigeo
dc.contributor.editorSkala, Václav
dc.date.accessioned2020-07-24T07:00:36Z-
dc.date.available2020-07-24T07:00:36Z-
dc.date.issued2020
dc.identifier.citationJournal of WSCG. 2020, vol. 28, no. 1-2, p. 79-88.en
dc.identifier.issn1213-6972 (print)
dc.identifier.issn1213-6980 (CD-ROM)
dc.identifier.issn1213-6964 (on-line)
dc.identifier.urihttp://wscg.zcu.cz/WSCG2020/2020-J_WSCG-1-2.pdf
dc.identifier.urihttp://hdl.handle.net/11025/38428
dc.format10 s.cs
dc.format.mimetypeapplication/pdf
dc.language.isoenen
dc.publisherVáclav Skala - UNION Agencycs
dc.relation.ispartofseriesJournal of WSCGen
dc.rights© Václav Skala - UNION Agencycs
dc.subjectodhad hloubkycs
dc.subjectporozumění scéněcs
dc.subjectrozšiřování datcs
dc.subject360 obrázkůcs
dc.titleForeground-aware Dense Depth Estimation for 360 Imagesen
dc.typečlánekcs
dc.typearticleen
dc.rights.accessopenAccessen
dc.type.versionpublishedVersionen
dc.description.abstract-translatedWith 360 imaging devices becoming widely accessible, omnidirectional content has gained popularity in multiple fields. The ability to estimate depth from a single omnidirectional image can benefit applications such as robotics navigation and virtual reality. However, existing depth estimation approaches produce sub-optimal results on real-world omnidirectional images with dynamic foreground objects. On the one hand, capture-based methods cannot obtain the foreground due to the limitations of the scanning and stitching schemes. On the other hand, it is challenging for synthesis-based methods to generate highly-realistic virtual foreground objects that are comparable to the real-world ones. In this paper, we propose to augment datasets with realistic foreground objects using an image-based approach, which produces a foreground-aware photorealistic dataset for machine learning algorithms. By exploiting a novel scale-invariant RGB-D orrespondence in the spherical domain, we repurpose abundant non-omnidirectional datasets to include realistic foreground objects with correct distortions. We further propose a novel auxiliary deep neural network to estimate both the depth of the omnidirectional images and the mask of the foreground objects, where the two tasks facilitate each other. A new local depth loss considers small regions of interests and ensures that their depth estimations are not smoothed out during the global gradient’s optimization. We demonstrate the system using human as the foreground due to its complexity and contextual importance, while the framework can be generalized to any other foreground objects. Experimental results demonstrate more consistent global estimations and more accurate local estimations compared with state-of-the-arts.en
dc.subject.translateddepth estimationen
dc.subject.translatedscene understandingen
dc.subject.translateddata augmentationen
dc.subject.translated360 imagesen
dc.identifier.doihttps://doi.org/10.24132/JWSCG.2020.28.10
dc.type.statusPeer-revieweden
Vyskytuje se v kolekcích:Volume 28, Number 1-2 (2020)

Soubory připojené k záznamu:
Soubor Popis VelikostFormát 
Feng.pdfPlný text10,45 MBAdobe PDFZobrazit/otevřít
Zobrazit minimální záznam Zobrazit statistiky


Použijte tento identifikátor k citaci nebo jako odkaz na tento záznam: http://hdl.handle.net/11025/38428

Všechny záznamy v DSpace jsou chráněny autorskými právy, všechna práva vyhrazena.

hledání
navigace
  1. DSpace at University of West Bohemia
  2. Publikace ZČU / UWB Publications
  3. Časopisy ZČU / UWB Journals
  4. Journal of WSCG
  5. Volume 28 (2020)
  6. Volume 28, Number 1-2 (2020)