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

Video salient object detection aims to find the most visually distinctiveobjects in a video. To explore the temporal dependencies, existing methodsusually resort to recurrent neural networks or optical flow. However, theseapproaches require high computational cost, and tend to accumulate inaccuraciesover time. In this paper, we propose a network with attention modules to learncontrastive features for video salient object detection without the highcomputational temporal modeling techniques. We develop a non-localself-attention scheme to capture the global information in the video frame. Aco-attention formulation is utilized to combine the low-level and high-levelfeatures. We further apply the contrastive learning to improve the featurerepresentations, where foreground region pairs from the same video are pulledtogether, and foreground-background region pairs are pushed away in the latentspace. The intra-frame contrastive loss helps separate the foreground andbackground features, and the inter-frame contrastive loss improves the temporalconsistency. We conduct extensive experiments on several benchmark datasets forvideo salient object detection and unsupervised video object segmentation, andshow that the proposed method requires less computation, and performs favorablyagainst the state-of-the-art approaches.

Video Salient Object Detection via Contrastive Features and Attention Modules