Skip to content

Latest commit

 

History

History
12 lines (8 loc) · 430 Bytes

README.md

File metadata and controls

12 lines (8 loc) · 430 Bytes

Hierarchical Question-Imagee Co-Attention for Visual Question Answering

This is MXNet implementation of Hierarchical Question-Image Co-Attention for Visual Question Answering(https://arxiv.org/abs/1606.00061)

Train

python train.py --cfg cfgs/hierarchical.yaml

Reference

[1]HieCoAttenVQA
[2]vqa-sva