INSTRUCTIONS

Dataset Annotations (Captions) Download

These are the steps to set up the dataset:-

1. Just use the images of the dataset given in the DSMLP cluster in /datasets/COCO-2015/
2. Create a sub-directory in the project root named datasets/
3. Create another sub-directory within it named as COCO/
4. For the captions, download the annotations from the MS COCO website.
5. Download the training set annotations as a zip file from here
6. Inflate inside the ./datasets/COCO/
7. Similary, download the zip file for 2014 testing image information from here and follow step 6.
8. Similarly, download the zip file for the 2015 testing image information from here and follow step 6.

Image loader

These are the steps to get your COCO dataset image loader up and running :-

1. Clone this repo recursively. To do this, run

cd src/
git submodule update --init --recursive

2. Build the submodule by running the following

cd src/cocoapi/PythonAPI/
make

3. Additionally symlink the pycocotools in the cocoapi's PythonAPI directory into src/ This can be done by the following

cd src/
ln -s ./cocoapi/PythonAPI/pycocotools ./