Repository for IFT 6266 H2017 project.

Steps to run:

1. Get datasets from : http://lisaweb.iro.umontreal.ca/transfert/lisa/datasets/mscoco_inpaiting/

2. Extract the inpainting directory. cd into the inpainting directory and create the following directories:

   1. pkls,
   2. models,
   3. predictions,
   4. word2vec,

3. Download the google trained gensim model binaries from: https://drive.google.com/file/d/0B7XkCwpI5KDYNlNUTTlSS21pQmM/edit into the word2vec directory

4. Install all required dependencies.

   1. numpy,
   2. scipy,
   3. sklearn,
   4. keras,
   5. theano,
   6. openCV,
   7. tqdm

5. In the full_script.py, change the inpainting_root variable to the folder where you extracted the data.

6. Run full_script.py