/imageqa-gen

A question generator described in paper "Exploring Model and Data for Image Question Answering"

Primary LanguagePythonMIT LicenseMIT

A Question Generator

Usage

python question_generator.py -parser_path {Stanford parser path} \
                             -sentence {Single sentence} \
                             -list {List of sentences file} \
                             -parsed_file {Parsed file} \
                             -output {Output file}

Prerequisites

  1. You need to download NLTK WordNet package.
python
import nltk
nltk.download()
d
wordnet

  1. You need to download Stanford Parser at http://nlp.stanford.edu/software/lex-parser.shtml#Download

  2. Extract the zip into a folder and remember the path

  3. You need to copy lexparser_sentence.sh into the Stanford Parser folder.

cp lexparser_sentence.sh stanford-parser/lexparser_sentence.sh

Examples

Run a single sentence

python question_generator.py -sentence "A man is riding a horse"

Run a list of sentences

  • Provide a file with each line in the file to be a sentence.
  • Output is a pickle file, storing a list.
  • Each element in the list is a tuple of five fields:
    1. Original sentence ID (0-based)
    2. Original sentence
    3. Generated question
    4. Answer to the generated question
    5. Type of the generated question
python question_generator.py -list sentences.txt -output questions.pkl

Run a pre-parsed file

Run stanford parser to pre-compute the parse trees.

lexparser.sh sentences.txt > sentences_parsed.txt

python question_generator.py -parsed_file sentences_parsed.txt \
                             -output questions.pkl

Reference

Exploring Models and Data for Image Question Answering. Mengye Ren, Ryan Kiros, Richard Zemel. NIPS, 2015.

@inproceedings{ren2015imageqa,
  title={Exploring Models and Data for Image Question Answering},
  author={Mengye Ren and Ryan Kiros and Richard Zemel},
  booktitle={NIPS},
  year={2015}
}