Global Regular Network for Writer Identification

TL;DR

This paper introduces a global regular network (GRN) that combines global page features and local word features for improved writer identification, achieving state-of-the-art accuracy efficiently.

Contribution

The paper proposes a novel GRN architecture with dual branches and residual attention for comprehensive handwriting feature extraction, outperforming existing methods.

Findings

Achieved 99.98% top-1 accuracy on CVL dataset

Reduced training time and network parameters compared to previous models

Demonstrated effectiveness of global and local feature integration

Abstract

Writer identification has practical applications for forgery detection and forensic science. Most models based on deep neural networks extract features from character image or sub-regions in character image, which ignoring features contained in page-region image. Our proposed global regular network (GRN) pays attention to these features. GRN network consists of two branches: one branch takes page handwriting as input to extract global features, and the other takes word handwriting as input to extract local features. Global features and local features merge in a global residual way to form overall features of the handwriting. The proposed GRN has two attributions: one is adding a branch to extract features contained in page; the other is using residual attention network to extract local feature. Experiments demonstrate the effectiveness of both strategies. On CVL dataset, our models achieve impressive 99.98% top-1 accuracy and 100% top-5 accuracy with shorter training time and fewer network parameters, which exceeded the state-of-the-art structure. The experiment shows the powerful ability of the network in the field of writer identification. The source code is available at https://github.com/wangshiyu001/GRN.