Learning ordinal constraint binary codes for fast similarity search

作者：

Highlights：

• We, for the first time, conceptualize the feature-level ordinal-preserving hashing for large-scale image retrieval.

• A discriminative ordinal-preserving latent graph hashing framework is designed to generate high-quality hash codes.

• Our method can capture the intrinsic latent features and ordinal-constraint local information into the learned hash codes.

• Our method jointly exploits latent space construction and ordinal graph learning in discriminative learning.

摘要

•We, for the first time, conceptualize the feature-level ordinal-preserving hashing for large-scale image retrieval.•A discriminative ordinal-preserving latent graph hashing framework is designed to generate high-quality hash codes.•Our method can capture the intrinsic latent features and ordinal-constraint local information into the learned hash codes.•Our method jointly exploits latent space construction and ordinal graph learning in discriminative learning.

论文关键词：Ordinal graph learning,Hashing-based image retrieval,Semantic-preserving hashing,Similarity comparison,Compact code learning

论文评审过程：Received 30 October 2021, Revised 6 February 2022, Accepted 24 February 2022, Available online 21 March 2022, Version of Record 21 March 2022.

论文官网地址：https://doi.org/10.1016/j.ipm.2022.102919