Adversarial Learning With Multi-Modal Attention for Visual Question Answering | IEEE Journals & Magazine | IEEE Xplore