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

Vector font synthesis is a challenging and ongoing problem in the fields ofComputer Vision and Computer Graphics. The recently-proposed DeepVecFontachieved state-of-the-art performance by exploiting information of both theimage and sequence modalities of vector fonts. However, it has limitedcapability for handling long sequence data and heavily relies on animage-guided outline refinement post-processing. Thus, vector glyphssynthesized by DeepVecFont still often contain some distortions and artifactsand cannot rival human-designed results. To address the above problems, thispaper proposes an enhanced version of DeepVecFont mainly by making thefollowing three novel technical contributions. First, we adopt Transformersinstead of RNNs to process sequential data and design a relaxationrepresentation for vector outlines, markedly improving the model's capabilityand stability of synthesizing long and complex outlines. Second, we propose tosample auxiliary points in addition to control points to precisely align thegenerated and target B\'ezier curves or lines. Finally, to alleviate erroraccumulation in the sequential generation process, we develop a context-basedself-refinement module based on another Transformer-based decoder to removeartifacts in the initially synthesized glyphs. Both qualitative andquantitative results demonstrate that the proposed method effectively resolvesthose intrinsic problems of the original DeepVecFont and outperforms existingapproaches in generating English and Chinese vector fonts with complicatedstructures and diverse styles.

DeepVecFont-v2: Exploiting Transformers to Synthesize Vector Fonts with Higher Quality