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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

The use of RGB-D information for salient object detection has beenextensively explored in recent years. However, relatively few efforts have beenput towards modeling salient object detection in real-world human activityscenes with RGBD. In this work, we fill the gap by making the followingcontributions to RGB-D salient object detection. (1) We carefully collect a newSIP (salient person) dataset, which consists of ~1K high-resolution images thatcover diverse real-world scenes from various viewpoints, poses, occlusions,illuminations, and backgrounds. (2) We conduct a large-scale (and, so far, themost comprehensive) benchmark comparing contemporary methods, which has longbeen missing in the field and can serve as a baseline for future research. Wesystematically summarize 32 popular models and evaluate 18 parts of 32 modelson seven datasets containing a total of about 97K images. (3) We propose asimple general architecture, called Deep Depth-Depurator Network (D3Net). Itconsists of a depth depurator unit (DDU) and a three-stream feature learningmodule (FLM), which performs low-quality depth map filtering and cross-modalfeature learning respectively. These components form a nested structure and areelaborately designed to be learned jointly. D3Net exceeds the performance ofany prior contenders across all five metrics under consideration, thus servingas a strong model to advance research in this field. We also demonstrate thatD3Net can be used to efficiently extract salient object masks from real scenes,enabling effective background changing application with a speed of 65fps on asingle GPU. All the saliency maps, our new SIP dataset, the D3Net model, andthe evaluation tools are publicly available athttps://github.com/DengPingFan/D3NetBenchmark.

Rethinking RGB-D Salient Object Detection: Models, Data Sets, and Large-Scale Benchmarks