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

Chinese characters carry a wealth of morphological and semantic information;therefore, the semantic enhancement of the morphology of Chinese characters hasdrawn significant attention. The previous methods were intended to directlyextract information from a whole Chinese character image, which usually cannotcapture both global and local information simultaneously. In this paper, wedevelop a stroke-based autoencoder(SAE), to model the sophisticated morphologyof Chinese characters with the self-supervised method. Following its canonicalwriting order, we first represent a Chinese character as a series of strokeimages with a fixed writing order, and then our SAE model is trained toreconstruct this stroke image sequence. This pre-trained SAE model can predictthe stroke image series for unseen characters, as long as their strokes orradicals appeared in the training set. We have designed two contrasting SAEarchitectures on different forms of stroke images. One is fine-tuned onexisting stroke-based method for zero-shot recognition of handwritten Chinesecharacters, and the other is applied to enrich the Chinese word embeddings fromtheir morphological features. The experimental results validate that afterpre-training, our SAE architecture outperforms other existing methods inzero-shot recognition and enhances the representation of Chinese characterswith their abundant morphological and semantic information.

Stroke-Based Autoencoders: Self-Supervised Learners for Efficient Zero-Shot Chinese Character Recognition