Modeling visual and word-conditional semantic attention for image captioning

作者：

Highlights：

•

• A new dual temporal model is proposed for image captioning.

• Word-conditional semantic attention is proposed for functional-words

• generation.

• A self-balancing model is exploited to balance the visual and semantic attention.

摘要

•A new dual temporal model is proposed for image captioning.•Word-conditional semantic attention is proposed for functional-words•generation.•A self-balancing model is exploited to balance the visual and semantic attention.

论文关键词：Image captioning,Word-conditional semantic attention,Visual attention,Attention variation

论文评审过程：Received 1 December 2017, Revised 28 May 2018, Accepted 4 June 2018, Available online 15 June 2018, Version of Record 21 June 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.06.002