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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Omni-directional images have been used in wide range of applications. For theapplications, it would be useful to estimate saliency maps representingprobability distributions of gazing points with a head-mounted display, todetect important regions in the omni-directional images. This paper proposes anovel saliency-map estimation model for the omni-directional images byextracting overlapping 2-dimensional (2D) plane images from omni-directionalimages at various directions and angles of view. While 2D saliency maps tend tohave high probability at the center of images (center bias), thehigh-probability region appears at horizontal directions in omni-directionalsaliency maps when a head-mounted display is used (equator bias). Therefore,the 2D saliency model with a center-bias layer was fine-tuned with anomni-directional dataset by replacing the center-bias layer to an equator-biaslayer conditioned on the elevation angle for the extraction of the 2D planeimage. The limited availability of omni-directional images in saliency datasetscan be compensated by using the well-established 2D saliency model pretrainedby a large number of training images with the ground truth of 2D saliency maps.In addition, this paper proposes a multi-scale estimation method by extracting2D images in multiple angles of view to detect objects of various sizes withvariable receptive fields. The saliency maps estimated from the multiple anglesof view were integrated by using pixel-wise attention weights calculated in anintegration layer for weighting the optimal scale to each object. The proposedmethod was evaluated using a publicly available dataset with evaluation metricsfor omni-directional saliency maps. It was confirmed that the accuracy of thesaliency maps was improved by the proposed method.

Multi-Scale Estimation for Omni-Directional Saliency Maps Using Learnable Equator Bias