Image-Captioning-Project

For Model 1 (CNN+LSTM)

Download the MSCOCO dataset from the below URL

http://images.cocodataset.org/zips/train2014.zip

Download Annotations File from the below URL

http://images.cocodataset.org/annotations/annotations_trainval2014.zip

Extract the zip files and place it in google drive

There are 5 files which should be executed cell by cell in following order

1) read_and_preprocess_data.ipynb

2) vocabulary_generator.ipynb

3) feature_extraction.ipynb

4) training_model.ipynb

5) testing_model.ipynb

For Model 2(With Visual attention)

Code need to be executed in Google Colab with GPU selected as runtime

Download the MSCOCO dataset from the below URL

http://images.cocodataset.org/zips/train2014.zip

Download Annotations File from the below URL

http://images.cocodataset.org/annotations/annotations_trainval2014.zip

Extract the zip files and place it in google drive

Place the Model2_VisualAttention.ipynb in the google drive

Execute all the cells of the file sequentially