Deep Co-Image-Label Hashing for Multi-Label Image Retrieval

Xiaobo Shen,Guohua Dong,Yuhui Zheng,Long Lan,Ivor W. Tsang,Quan-Sen Sun,Ivor Tsang
DOI: https://doi.org/10.1109/tmm.2021.3119868
IF: 7.3
2022-01-01
IEEE Transactions on Multimedia
Abstract:Deep supervised hashing has greatly improved retrieval performance with the powerful learning capability of deep neural network. In multi-label image retrieval, existing deep hashing simply indicates whether two images are similar by constructing a similarity matrix. However, it ignores the dependency among multiple labels that has been shown important in multi-label application. To fulfill this gap, this paper proposes Deep Co-Image-Label Hashing (DCILH) to discover label dependency. Specifically, DCILH regards image and label as two views, and maps the two views into a common deep Hamming space. DCILH proposes to learn prototype for each label, and preserve similarity among images, labels, and prototypes. To exploit label dependency, DCILH further employs the label-correlation aware loss on the predicted labels, such that predicted output on positive label is enforced to be larger than that on negative label. Extensive experiments on several multi-label benchmarks demonstrate the proposed DCILH outperforms state-of-the-art deep supervised hashing on large-scale multi-label image retrieval.
computer science, information systems,telecommunications, software engineering
What problem does this paper attempt to address?