Skip to content

Chetan45s/Image-Caption-Bot

Repository files navigation

Image-Caption-Bot

Description of the project

Created a Deep Learning Model that Caption the Image with good accuracy using RNN,CNN,word2vec and Keras.

In this project we worked on the flicker8k dataset and tried to train a deep learning model that can caption the image by itself. Further more we hosted the trained model using django and connect the model to the bot that can take a image as a input and output the image with captions on it.

Some sample captions that are generated

1 2 3

Tech

Image Caption Bot uses following technologies:

  • [HTML CSS] - For good GUI.
  • [Python] - All the backend and machine learning code is writtern in Python.
  • [Tensorflow] - To server keras a backend
  • [Django] - For backend services

Installation

Image Caption Bot requires Python v3+ to run.

git clone https://github.com/Chetan45s/Image-Caption-Bot.git

cd Image-Caption-Bot
pip install virtualenv

mkvirtualenv image_bot

workon image_bot

pip install -r requirement.txt (This may take some time as we required to install tensorflow)

python manage.py runserver

Now, switch to https://locahost:8000

It will look something like this

alt text alt text