Amazing Bahasa Indonesia Image Captioning

Citation

This repository is implementation of this paper

Please cite using this bib

@INPROCEEDINGS{8835370,
  author={A. A. {Nugraha} and A. {Arifianto} and  {Suyanto}},
  booktitle={2019 7th International Conference on Information and Communication Technology (ICoICT)}, 
  title={Generating Image Description on Indonesian Language using Convolutional Neural Network and Gated Recurrent Unit}, 
  year={2019},
  volume={},
  number={},
  pages={1-6},
  doi={10.1109/ICoICT.2019.8835370}}

or using this

A. A. Nugraha, A. Arifianto and Suyanto, "Generating Image Description on Indonesian Language using Convolutional Neural Network and Gated Recurrent Unit," 2019 7th International Conference on Information and Communication Technology (ICoICT), Kuala Lumpur, Malaysia, 2019, pp. 1-6, doi: 10.1109/ICoICT.2019.8835370.

Prerequisites

This tutorial only applies to UNIX like systems. If your system not an UNIX type, please search the tutorial based on your system.

What you need to have

• Miniconda with Python 3
• Docker

Install Miniconda with Python 3

wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

Create an environment and install the library

conda env create
source activate image-captioning

Download the Dataset

Create dataset folder

mkdir dataset
cd dataset

Flickr8k Dataset

M. Hodosh, P. Young and J. Hockenmaier (2013) "Framing Image Description as a Ranking Task: Data, Models and Evaluation Metrics", Journal of Artifical Intellegence Research, Volume 47, pages 853-899 http://www.jair.org/papers/paper3994.html

Download Images and Caption

wget http://nlp.cs.illinois.edu/HockenmaierGroup/Framing_Image_Description/Flickr8k_Dataset.zip
wget http://nlp.cs.illinois.edu/HockenmaierGroup/Framing_Image_Description/Flickr8k_text.zip

Unzip

unzip Flickr8k_Dataset.zip
unzip Flickr8k_text.zip

Tutorial

Train the model

python -m src.train -m <your-model-type>

Model type:

1: train using image caption sentence modeler
2: train using image caption single word modeler

Serving the model

python -m src.train -m <your-model-type>

Model type:

1: serving using image caption sentence modeler
2: serving using image caption single word modeler

Predict Image

python -m src.predict -m <your-model-type> -p <your-image-path> [OPTION]

Model type

1: predict using image caption sentece modeler
2: predict using image caption single word modeler

Option

• --show-image : show captioned image with matplotlib

Predict Video frame by frame

python -m src.run_video <your-video-path>

If video path is empty, it will use your webcam as video stream

Predict CCTV frame by frame

python -m src.run_cctv

Results

Image Captioning

CCTV Captioning

Authors

• Aditya Alif Nugraha - Aditya Alif Nugraha

Resources

• Convolutional Neural Network - deeplearning.ai
• Convolutional Neural Network - Stanford University