SDU-MIMA / SHDCH

This is a simplified demo for the paper: Supervised Hierarchical Deep Hashing for Cross-Modal Retrieval

our method

Demo on FashionVC dataset.

Description

This is a simplified demo for the paper：Supervised Hierarchical Deep Hashing for Cross-Modal Retrieval, including:
load_data.py: function to load data.
img_net.py: net structure for image modality.
txt_net.py: net structure for text modality.
demo.py: a demo for our method on FashionVC dataset.
other: tools to evaluate the preformance of the method.

Version

• Python 3.5
• Tensorflow 1.13.1

How-to run

1. Download the FashionVC dataset, and put it into ./DataSet
   you can download dataset from pan.baidu.com
   link: https://pan.baidu.com/s/1VZwdU8MhWkvVmpMjrFJktw
   password: sreu
2. Download the Ssense dataset, and put it into ./DataSet
   you can download dataset from pan.baidu.com
   link: https://pan.baidu.com/s/1RZsSZY5pY2GSAQEu5ciqAw
   password: qwmq
3. Download the imagenet-vgg-f.mat, and put it into ./DataSet
   you can download it from pan.baidu.com
   link: https://pan.baidu.com/s/1jM9ZPXGLIykw4SLzk3_3Wg
   password: 9i4a
4. run demo.py