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

The extraction of text in high quality is essential for text-based documentanalysis tasks like Document Classification or Named Entity Recognition.Unfortunately, this is not always ensured, as poor scan quality and theresulting artifacts lead to errors in the Optical Character Recognition (OCR)process. Current approaches using Convolutional Neural Networks show promisingresults for background removal tasks but fail correcting artifacts likepixelation or compression errors. For general images, Transformer backbones aregetting integrated more frequently in well-known neural network structures fordenoising tasks. In this work, a modified UNet structure using a SwinTransformer backbone is presented to remove typical artifacts in scanneddocuments. Multi-headed cross-attention skip connections are used to moreselectively learn features in respective levels of abstraction. The performanceof this approach is examined regarding compression errors, pixelation andrandom noise. An improvement in text extraction quality with a reduced errorrate of up to 53.9% on the synthetic data is archived. The pretrainedbase-model can be easily adapted to new artifacts. The cross-attention skipconnections allow to integrate textual information extracted from the encoderor in form of commands to more selectively control the models outcome. Thelatter is shown by means of an example application.

Transformer-Based UNet with Multi-Headed Cross-Attention Skip Connections to Eliminate Artifacts in Scanned Documents