Supervised Deep Hashing for Hierarchical Labeled Data


  • Dan Wang Beijing Institute of Technology
  • Heyan Huang Beijing Institute of Technology
  • Chi Lu Beijing Institute of Technology
  • Bo-Si Feng Beijing Institute of Technology
  • Guihua Wen South China University of Technology
  • Liqiang Nie Shandong University
  • Xian-Ling Mao Beijing Institute of Technology



Deep Hashing, Hierarchical Labeled Data, Supervised Learning


Recently, hashing methods have been widely used in large-scale image retrieval. However, most existing supervised hashing methods do not consider the hierarchical relation of labels,which means that they ignored the rich semantic information stored in the hierarchy. Moreover, most of previous works treat each bit in a hash code equally, which does not meet the scenario of hierarchical labeled data. To tackle the aforementioned problems, in this paper, we propose a novel deep hashing method, called supervised hierarchical deep hashing (SHDH), to perform hash code learning for hierarchical labeled data. Specifically, we define a novel similarity formula for hierarchical labeled data by weighting each level, and design a deep neural network to obtain a hash code for each data point. Extensive experiments on two real-world public datasets show that the proposed method outperforms the state-of-the-art baselines in the image retrieval task.




How to Cite

Wang, D., Huang, H., Lu, C., Feng, B.-S., Wen, G., Nie, L., & Mao, X.-L. (2018). Supervised Deep Hashing for Hierarchical Labeled Data. Proceedings of the AAAI Conference on Artificial Intelligence, 32(1).