Skip to content
No description, website, or topics provided.
Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data Initial model file upload Oct 7, 2019
model Initial code file upload Oct 7, 2019
w2v_models/gensim_models_splits Initial model file upload Oct 7, 2019
README.md Update README.md Oct 9, 2019
main.py Initial code file upload Oct 7, 2019

README.md

MovieQAWithoutMovies

Ranked 1st on MovieQA leaderboard on 4 out of 5 categories.

Wiki

We have provided the processed data to run the code and get different results as mentioned in the paper. The accompanying paper can be found at the project page. Follow the below steps to run our code.

Setup the repository

  1. Clone this repo - git clone https://github.com/BhavanJ/MovieQAWithoutMovies
  2. Download the contents of data and w2v folders from here. These are the processed data.
  3. Download the file GoogleNews-vectors-negative300.bin.gz from here and save it in w2v_model/Glove

Train and replicate the results

python main.py

TODO: explain different settings

Paper

Bhavan Jasani, Rohit Girdhar, Deva Ramanan, "Are we asking the right questions in MovieQA?", ICCVW, 2019.

@inproceedings{BJ_ICCV_2019,
  author    = {Bhavan Jasani, Rohit Girdhar and Deva Ramanan},
  title     = {Are we asking the right questions in MovieQA?},
  booktitle = {ICCVW},
  year      = {2019},
}

This code is built upon Layered Memory Network. We would like to thank them for providing some of their processed data.

You can’t perform that action at this time.