/Let-me-see

Let me see: an encoder-decoder image captioning

Primary LanguagePython

Let me see

Let me see: an encoder-decoder image captioning

Credits

This repository is based on previous work of:

  • Vinyals, O. T. (2015). Show and Tell: A neural image Caption generator. arXiv, 1411.4555.
  • Tanti, M. G. (2018). Where to put the Image in an Image Caption Generator. arXiv:, 1703.09137.

Author

Jose Ignacio Bengoechea Isasa, ignacio.bengis@gmail.com

Quickstart

  • Language: Python 3.7.
  • Requires tensorflow, keras, pillow, numpy, pandas.

Assuming git, python and pip installed:

 git clone https://github.com/Bengis/Let-me-see
 cd Let-me-see/code
 python preprocessing.py
 python encoder.py
 python train.py

Dataset

The dataset of this software is Flickr8K. You can request here to download the dataset.

About this software

  • This software is part of the final proyect of the Master of Data Science
  • Master of Data Science.
  • Universitat Oberta of Catalunya.
  • Tutored by: Anna Bosch RuĂ©