Research Article

JGRCAN: A Visual Question Answering Co-Attention Network via Joint Grid-Region Features

Figure 7

Some samples from VQA-v2.