-
This is the repo of RHA model. This paper has been published on ICMR 2021, you can find this paper here
-
The code is modified from the baseline STAGE model. Some comments or code may be hard to read, since I recently have no time for code cleaning. If you have any questions, please submit a issue.
If this work is helpful, please cite as
@article{li2021relation,
title={Relation-aware Hierarchical Attention Framework for Video Question Answering},
author={Li, Fangtao and Bai, Ting and Cao, Chenyu and Liu, Zihe and Yan, Chenghao and Wu, Bin},
journal={arXiv preprint arXiv:2105.06160},
year={2021}
}