Cross_Modal_GCN

Semi-supervised Graph Convolutional Hashing Network For Large-Scale Cross-Modal Retrieval (Shen, Zhai et al., ICIP 2020)

Here we provide the implementation of our model in TensorFlow, along with all experimental datasets. The repository is organised as follows:

data/ contains the necessary dataset files. Here, we have three datasets including MIRFLICKR-25K, NUS-WIDE-10K and Wiki, which can be downloaded from pan.baidu.com: link: https://pan.baidu.com/s/1DlIxCvT_3vRKphMydnljVQ code: b6be.
train_semi_nuswide.py execute a full training run on NUS-WIDE-10K dataset.
train_semi_wiki.py execute a full training run on Wiki dataset.

The script has been tested running under Python 2.7, with the following packages installed (along with their dependencies):

In addition, CUDA 9.0 and cuDNN 7 have been used.

Name		Name	Last commit message	Last commit date
Latest commit History 16 Commits
README.md		README.md
SGCH.pdf		SGCH.pdf
__init__.py		__init__.py
flip_gradient.py		flip_gradient.py
framework.png		framework.png
inits.py		inits.py
layers.py		layers.py
metrics.py		metrics.py
models.py		models.py
train_semi_nuswide.py		train_semi_nuswide.py
train_semi_wiki.py		train_semi_wiki.py
utils.py		utils.py

Provide feedback