/DCDH

Source code for Deep Collaborative Discrete Hashing with Semantic-Invariant Structure -- PyTorch Implementation

Primary LanguagePythonMIT LicenseMIT

Deep Collaborative Discrete Hashing with Semantic-Invariant Structure -- PyTorch Implementation

"Deep Collaborative Discrete Hashing with Semantic-Invariant Structure", Published in The 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), 2019.

Implementation platform:

  • Python 3.6
  • Pytorch 1.1

Datasets

We also release the train/query split for the following datasets.

  • CIFAR-10
  • MIRFlickr-25k

Training

The general command for training is

  • python3 DCDH_Flickr.py
  • python3 DCDH_CIFAR_10.py

If you find the code is useful, please cite the following references:

[1] Z. Wang*, Z. Zhang*, Y. Luo, Z. Huang, Deep Collaborative Discrete Hashing with Semantic-Invariant Structure, in Proc. of The 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), Paris, France, 2019. Paper link

[2] Z. Wang*, Z. Zhang*, Y. Luo, Z. Huang, Heng Tao Shen, Deep Collaborative Discrete Hashing with Semantic-Invariant Structure, IEEE Transactions on Multimedia, 2020, Journal Extension of [1]. Paper Link