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

Automatic recognition of the historical letters (XI-XVIII centuries) carvedon the stoned walls of St.Sophia cathedral in Kyiv (Ukraine) was demonstratedby means of capsule deep learning neural network. It was applied to the imagedataset of the carved Glagolitic and Cyrillic letters (CGCL), which wasassembled and pre-processed recently for recognition and prediction by machinelearning methods(https://www.kaggle.com/yoctoman/graffiti-st-sophia-cathedral-kyiv). CGCLdataset contains >4000 images for glyphs of 34 letters which are hardlyrecognized by experts even in contrast to notMNIST dataset with the betterimages of 10 letters taken from different fonts. Despite the much worse qualityof CGCL dataset and extremely low number of samples (in comparison to notMNISTdataset) the capsule network model demonstrated much better results than thepreviously used convolutional neural network (CNN). The validation accuracy(and validation loss) was higher (lower) for capsule network model than for CNNwithout data augmentation even. The area under curve (AUC) values for receiveroperating characteristic (ROC) were also higher for the capsule network modelthan for CNN model: 0.88-0.93 (capsule network) and 0.50 (CNN) without dataaugmentation, 0.91-0.95 (capsule network) and 0.51 (CNN) with lossless dataaugmentation, and similar results of 0.91-0.93 (capsule network) and 0.9 (CNN)in the regime of lossless data augmentation only. The confusion matrixes weremuch better for capsule network than for CNN model and gave the much lower typeI (false positive) and type II (false negative) values in all three regimes ofdata augmentation. These results supports the previous claims that capsule-likenetworks allow to reduce error rates not only on MNIST digit dataset, but onthe other notMNIST letter dataset and the more complex CGCL handwritinggraffiti letter dataset also.

Capsule Deep Neural Network for Recognition of Historical Graffiti Handwriting