Skip to content

Video Question Answering via Hierarchical Spatio-Temporal Attention Networks

Notifications You must be signed in to change notification settings

yqf-oo/videoqa-stan

Repository files navigation

videoqa-stan

Source code for Video Question Answering via Hierarchical Spatio-Temporal Attention Networks based on arctic-captions and arctic-capgen-vid.

Dependencies

To run this code you will need:

  • Python 2.7
  • Theano
  • NLTK and WordNet Synsets

Reference

"Video Question Answering via Hierarchical Spatio-Temporal Attention Networks."
Zhou Zhao, Qifan Yang*, Deng Cai, Xiaofei He, Yueting Zhuang. To appear IJCAI 2017.

License

The code is released under a revised (3-clause) BSD License.

About

Video Question Answering via Hierarchical Spatio-Temporal Attention Networks

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages