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/

Zendy Plus

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

Zendy Tools

Zendy Open

To reduce the computational cost of attention, this paper proposes the semantic web-driven efficient self-attention (SWDESA) model. It stacks multiple hierarchical transformer modules and expands the window size to capture feature information at different scales. Additionally, SWDESA splits the feature map and compresses its spatial dimensions through linear mapping, reducing the computational burden. The model introduces separation and restoration self-attention to aggregate spatial and channel information, with the computational cost being linearly related to the window size. By utilizing semantic information to guide the separation and restoration self-attention in establishing long-range relationships, the model optimizes semantic perception and enhances information aggregation. Multi-branch parallel separable convolutions such as a feed-forward network are introduced to extract fine-grained textures of the main objects and macro features of the background selectively. Experimental results on the DIV2K dataset show that SWDESA model achieved a peak signal-to-noise ratio of 68.91 and a structural similarity index of 68.84, achieving the best performance.

Semantic Web-Driven Efficient Self-Attention Mechanism for High-Resolution Image Reconstruction