Multiple attentional aggregation network for handwritten Dongba character recognition

作者：

Highlights：

• A novel Dongba character image dataset is constructed.

• A novel framework for Dongba character recognition is proposed.

• Two feature extraction units are proposed for richer feature expression.

• Experimental results demonstrate the MAAN outperforms state-of-the-art works.

摘要

•A novel Dongba character image dataset is constructed.•A novel framework for Dongba character recognition is proposed.•Two feature extraction units are proposed for richer feature expression.•Experimental results demonstrate the MAAN outperforms state-of-the-art works.

论文关键词：Aggregation network,Attention mechanism,Multi-scale features,Dongba character recognition,Dongba character image dataset

论文评审过程：Received 12 July 2022, Revised 4 September 2022, Accepted 16 September 2022, Available online 22 September 2022, Version of Record 21 October 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118865