The following project takes in images and returns the caption for image
The purpose of this project is to help visually impaired in identifying objects and actions in an image.
The aim of this project is to construct a neural network architecture (CNN Encoder- RNN Decoder) which generates captions from images automatically.
Built using technologies of two subdomains of Machine Learning : Computer Vision and Natural Language Processing.
Trained on Flickr8k dataset and the network is tested on the novel images to give 82.4 percent accuracy.
The model gave the following results
