Cognitive multi-modal consistent hashing with flexible semantic transformation
作者：
Highlights：
• Discriminative cognitive multi-modal hashing is proposed to achieve multi-modal hash codes.
• Latent space construction, semantic transformation and cognitive feature learning are jointly considered.
• Collaborative multi-modal fusion module explores underlying common components from multi-source data.
• A cognitive self-paced learning policy conducts progressive feature aggregation.
摘要
•Discriminative cognitive multi-modal hashing is proposed to achieve multi-modal hash codes.•Latent space construction, semantic transformation and cognitive feature learning are jointly considered.•Collaborative multi-modal fusion module explores underlying common components from multi-source data.•A cognitive self-paced learning policy conducts progressive feature aggregation.
论文关键词：Social geo-media,Learning to hash,Semantic preserving,Discrete optimization,Similarity search
论文评审过程：Received 28 April 2021, Revised 19 August 2021, Accepted 29 August 2021, Available online 24 September 2021, Version of Record 24 September 2021.
论文官网地址：https://doi.org/10.1016/j.ipm.2021.102743
