Pytorch Implementation of Adaptive Attention Model for Image Captioning

Knowing When to Look: Adaptive Attention via A Visual Sentinel for Image Captioning [Paper] [Review]

Dataset Preparation

First we will need to download the MS-COCO dataset. So create a data folder and run the download bash script

mkdir data && ./download.sh

Afterwards, we should create the Karpathy split for training, validation and test.

python KarpathySplit.py

Then we can build the vocabulary by running

python build_vocab.py

The vocab.pkl should be saved in the data folder.

Now we will need to resize all the images in both train and val folder. Here I create a new folder under data, i.e., 'resized'. Then we may run resize.py to resize all images into 256 x 256. You may specify different locations inside resize.py

mkdir data/resized && python resize.py

After all images are resized. Now we can train our Adaptive Attention model with

python train.py