USCDataScience / img2text

Models, and associated helper code for GSOC 2017 project Tensorflow Image to Text in Apache Tika

img2text

This project contains a Tensorflow trained model that implements the GSOC 2017 project Tensorflow Image to Text in Apache Tika based on the paper Show and Tell: A Neural Image Caption Generator. The model is split into multiple parts, and the associated Docker image in Apache Tika puts the model back together during the docker process for use in Tika.

How to obtain the checkpoint file?

git clone https://github.com/USCDataScience/img2text.git
cd img2text
sudo chmod a+x download-1M_iters_ckpt.sh
./download-1M_iters_ckpt.sh

Questions, comments?

Send them to Chris A. Mattmann or Thejan Wijesinghe.

Contributors

• Chris A. Mattmann, JPL & USC
• Thejan Wijesinghe, University of Moratuwa

License

This project is licensed under the Apache License, version 2.0.