GR-RNN: Global-context residual recurrent neural networks for writer identification

摘要

This paper presents an end-to-end neural network system to identify writers through handwritten word images, which jointly integrates global-context information and a sequence of local fragment-based fea-tures. The global-context information is extracted from the tail of the neural network by a global aver-age pooling step. The sequence of local and fragment-based features is extracted from a low-level deep feature map which contains subtle information about the handwriting style. The spatial relationship be-tween the sequence of fragments is modeled by the recurrent neural network (RNN) to strengthen the discriminative ability of the local fragment features. We leverage the complementary information be-tween the global-context and local fragments, resulting in the proposed global-context residual recurrent neural network (GR-RNN) method. The proposed method is evaluated on four public data sets and ex-perimental results demonstrate that it can provide state-of-the-art performance. In addition, the neural networks trained on gray-scale images provide better results than neural networks trained on binarized and contour images, indicating that texture information plays an important role for writer identification. The source code is available: https://github.com/shengfly/writer-identification .