Code for paper "Deep Co-Attention Network for Multi-View Subspace Learning"
- Python 3.7
- Pytorch
- Tensorflow (For downloading mnist dataset)
- Download dataset from http://www.vision.caltech.edu/visipedia/CUB-200.html
- Download the mask-rcnn from https://github.com/matterport/Mask_RCNN, and install the required package.
- Run the following command to generate two views as well as the image segments:
python mask_rcnn.py
- Run the following command to get the hidden representation for the final training:
python pre-train_vgg.py
- Run the following command to train the model:
python main.py -g 0 -e 150 -hid 300 -d birds