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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In this work we present FreDSNet, a deep learning solution which obtainssemantic 3D understanding of indoor environments from single panoramas.Omnidirectional images reveal task-specific advantages when addressing sceneunderstanding problems due to the 360-degree contextual information about theentire environment they provide. However, the inherent characteristics of theomnidirectional images add additional problems to obtain an accurate detectionand segmentation of objects or a good depth estimation. To overcome theseproblems, we exploit convolutions in the frequential domain obtaining a widerreceptive field in each convolutional layer. These convolutions allow toleverage the whole context information from omnidirectional images. FreDSNet isthe first network that jointly provides monocular depth estimation and semanticsegmentation from a single panoramic image exploiting fast Fourierconvolutions. Our experiments show that FreDSNet has similar performance asspecific state of the art methods for semantic segmentation and depthestimation. FreDSNet code is publicly available inhttps://github.com/Sbrunoberenguel/FreDSNet

FreDSNet: Joint Monocular Depth and Semantic Segmentation with Fast Fourier Convolutions