VQA

Hierarchical Question-Image Co-Attention for Visual Question Answering

Train

Simple Baseline

python3 -m coatt.main --model simple

Coattention Network

python3 -m coatt.main --model coattention

The code is based on the following papers:

[1] VQA: Visual Question Answering (Agrawal et al, 2016): https://arxiv.org/pdf/1505.00468v6.pdf

[2] Simple Baseline for Visual Question Answering (Zhou et al, 2015): https://arxiv.org/pdf/1512.02167.pdf

[3] Hierarchical Question-Image Co-Attention for Visual Question Answering (Lu et al, 2017): https://arxiv.org/pdf/1606.00061.pdf