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

Recently deep convolutional neural networks have achieved significant successin salient object detection. However, existing state-of-the-art methods requirehigh-end GPUs to achieve real-time performance, which makes them hard to adaptto low-cost or portable devices. Although generic network architectures havebeen proposed to speed up inference on mobile devices, they are tailored to thetask of image classification or semantic segmentation, and struggle to captureintra-channel and inter-channel correlations that are essential for contrastmodeling in salient object detection. Motivated by the above observations, wedesign a new deep learning algorithm for fast salient object detection. Theproposed algorithm for the first time achieves competitive accuracy and highinference efficiency simultaneously with a single CPU thread. Specifically, wepropose a novel depthwise non-local moudule (DNL), which implicitly modelscontrast via harvesting intra-channel and inter-channel correlations in aself-attention manner. In addition, we introduce a depthwise non-local networkarchitecture that incorporates both depthwise non-local modules and invertedresidual blocks. Experimental results show that our proposed network attainsvery competitive accuracy on a wide range of salient object detection datasetswhile achieving state-of-the-art efficiency among all existing deep learningbased algorithms.

Depthwise Non-local Module for Fast Salient Object Detection Using a Single Thread