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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Purpose. Handwriting is one of the most frequently occurring patterns ineveryday life and with it come challenging applications such as handwritingrecognition (HWR), writer identification, and signature verification. Incontrast to offline HWR that only uses spatial information (i.e., images),online HWR (OnHWR) uses richer spatio-temporal information (i.e., trajectorydata or inertial data). While there exist many offline HWR datasets, there isonly little data available for the development of OnHWR methods on paper as itrequires hardware-integrated pens. Methods. This paper presents data andbenchmark models for real-time sequence-to-sequence (seq2seq) learning andsingle character-based recognition. Our data is recorded by a sensor-enhancedballpoint pen, yielding sensor data streams from triaxial accelerometers, agyroscope, a magnetometer and a force sensor at 100 Hz. We propose a variety ofdatasets including equations and words for both the writer-dependent andwriter-independent tasks. Our datasets allow a comparison between classicalOnHWR on tablets and on paper with sensor-enhanced pens. We provide anevaluation benchmark for seq2seq and single character-based HWR using recurrentand temporal convolutional networks and Transformers combined with aconnectionist temporal classification (CTC) loss and cross-entropy (CE) losses.Results. Our convolutional network combined with BiLSTMs outperformsTransformer-based architectures, is on par with InceptionTime forsequence-based classification tasks, and yields better results compared to 28state-of-the-art techniques. Time-series augmentation methods improve thesequence-based task, and we show that CE variants can improve the singleclassification task.

Benchmarking Online Sequence-to-Sequence and Character-based Handwriting Recognition from IMU-Enhanced Pens