Towards efficient unconstrained handwriting recognition using Dilated Temporal Convolution Network

作者：

Highlights：

• Effectiveness of TCN based handwriting recognition model for line images.

• Analysis of the model with and without data augmentation.

• Analysis of the model with different scales of the line images.

• Resource requirements, training and testing time comparison with RNN architectures.

摘要

•Effectiveness of TCN based handwriting recognition model for line images.•Analysis of the model with and without data augmentation.•Analysis of the model with different scales of the line images.•Resource requirements, training and testing time comparison with RNN architectures.

论文关键词：Dilated Temporal Convolution Network,Handwriting recognition,Document analysis

论文评审过程：Received 19 March 2020, Revised 10 September 2020, Accepted 11 September 2020, Available online 12 September 2020, Version of Record 15 September 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114004