/videoQA

Primary LanguagePython

EECS 598 project

Getting SceneGraphs.

  • The scenegraphs are got via a hacked library of CORENLP’s scenegraph generator. This will be in the local copy presented (bin/). However, due to limitations of git and the lack of clarity on our git repository’s library.

Installation

  • The below will take a lot of time as we need a lot of resources that comes along with the libraries. (nltk, networkx, spacy are huge libraries and their dependencies are also very large).
git clone https://github.com/AkashGanesan/video-context-transcription   
cd video-context-transcription    
module load python-dev/3.5.2 # Needs Python3
virtualenv  -p python3 env
source env/bin/activate
pip install --upgrade pip
pip install networkx nltk numpy pathlib easydict scipy spacy matplotlib
python -m spacy download en_core_web_lg
python -m nltk.downloader all
  • Once the above has been done, we have everything to analyze videos. You can run python search.py for a demo.
cd src/python/
python search.py
  • NOTE: The power plot code has been commented out as the flux machine doesn’t have tkinter.