To create a deep learning model which can explain the content of an image in the form of speech through caption generation with attention mechanism on Flickr8K dataset.