Learning joint relationship attention network for image captioning
作者：
Highlights：
• A joint relationship attention network is proposed to enhance image captioning.
• Joint relationship learning network is built to learn two types of relationships.
• Relationship gate is introduced to balance the contributions of two relationships.
• A feature fusion-based attention is used to boost the model performance.
• Various aspects of our proposed approach are extensively evaluated.
摘要
•A joint relationship attention network is proposed to enhance image captioning.•Joint relationship learning network is built to learn two types of relationships.•Relationship gate is introduced to balance the contributions of two relationships.•A feature fusion-based attention is used to boost the model performance.•Various aspects of our proposed approach are extensively evaluated.
论文关键词：Visual relationship,Feature relationship network,Joint relationship learning,Image captioning
论文评审过程：Received 29 January 2022, Revised 17 June 2022, Accepted 5 August 2022, Available online 19 August 2022, Version of Record 29 August 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.118474
